post_id
stringlengths 5
7
| domain
stringclasses 18
values | upvote_ratio
float64 0.5
1
| history
stringlengths 22
39.2k
| c_root_id_A
stringlengths 7
7
| c_root_id_B
stringlengths 7
7
| created_at_utc_A
int64 1.28B
1.67B
| created_at_utc_B
int64 1.28B
1.67B
| score_A
int64 2
43.5k
| score_B
int64 2
43.2k
| human_ref_A
stringlengths 0
10.7k
| human_ref_B
stringlengths 0
10.8k
| labels
int64 0
1
| seconds_difference
float64 0
145M
| score_ratio
float64 1
3.72k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivg05r0 | 1,667,840,514 | 1,667,845,336 | 97 | 111 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 4,822 | 1.14433 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivf1b60 | 1,667,845,336 | 1,667,831,428 | 111 | 84 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | 1 | 13,908 | 1.321429 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivf59r0 | 1,667,845,336 | 1,667,833,145 | 111 | 78 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | 1 | 12,191 | 1.423077 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | iveslwa | 1,667,845,336 | 1,667,827,288 | 111 | 74 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | I'm in religion, technically in 'theology' not religious studies. I train ministers and chaplains. I'm pretty sure God doesn't exist. More agnostic than atheist, but even if God does exist, I don't think They matter that much. I think all the time angsting over who God is or what God wants is a distraction from making a better society here and now. But I try to understand why others think it's so important. | 1 | 18,048 | 1.5 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivf4p3l | 1,667,845,336 | 1,667,832,901 | 111 | 68 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | 1 | 12,435 | 1.632353 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfgxyt | 1,667,845,336 | 1,667,837,919 | 111 | 61 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 1 | 7,417 | 1.819672 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfb7ap | 1,667,845,336 | 1,667,835,618 | 111 | 59 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | 1 | 9,718 | 1.881356 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | iveu9yt | 1,667,845,336 | 1,667,828,142 | 111 | 50 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | 1 | 17,194 | 2.22 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivesaad | 1,667,845,336 | 1,667,827,120 | 111 | 50 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 18,216 | 2.22 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfyz8k | 1,667,845,336 | 1,667,844,890 | 111 | 46 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 1 | 446 | 2.413043 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | ivg05r0 | 1,667,838,114 | 1,667,845,336 | 41 | 111 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 7,222 | 2.707317 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | iveud6r | 1,667,845,336 | 1,667,828,186 | 111 | 31 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | 1 | 17,150 | 3.580645 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfena9 | 1,667,845,336 | 1,667,837,000 | 111 | 29 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Neuroscience: There is no free will. | 1 | 8,336 | 3.827586 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivg05r0 | 1,667,838,362 | 1,667,845,336 | 36 | 111 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 6,974 | 3.083333 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfkric | ivg05r0 | 1,667,839,428 | 1,667,845,336 | 28 | 111 | We all talk a good game about null findings and replication, but it's thankless and minimally funded work. People kinda care, but only when somebody else does it. | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 5,908 | 3.964286 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfl10y | ivg05r0 | 1,667,839,532 | 1,667,845,336 | 29 | 111 | Sometimes I think that philosophy should turn all of the classic stuff --- everything from Socrates and Plato to (oh) Nietzsche and Russell --- over to history departments. I think this would be better for philosophy in the long run, but mostly I don't think we can be trusted with it. | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 5,804 | 3.827586 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf10s1 | ivg05r0 | 1,667,831,300 | 1,667,845,336 | 26 | 111 | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | 0 | 14,036 | 4.269231 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfd2kp | 1,667,845,336 | 1,667,836,370 | 111 | 24 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | 1 | 8,966 | 4.625 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg05r0 | ivfg4n4 | 1,667,845,336 | 1,667,837,592 | 111 | 28 | Analytical chemist in cannabis testing. Cannabinoids are not a miracle drug but we all have to pretend it might be to give the industry validity. Why can’t we just admit that most of us just like getting stoned and that’s okay. | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | 1 | 7,744 | 3.964286 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivf1b60 | 1,667,840,514 | 1,667,831,428 | 97 | 84 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | 1 | 9,086 | 1.154762 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivf59r0 | 1,667,840,514 | 1,667,833,145 | 97 | 78 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | 1 | 7,369 | 1.24359 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveslwa | ivfnk1y | 1,667,827,288 | 1,667,840,514 | 74 | 97 | I'm in religion, technically in 'theology' not religious studies. I train ministers and chaplains. I'm pretty sure God doesn't exist. More agnostic than atheist, but even if God does exist, I don't think They matter that much. I think all the time angsting over who God is or what God wants is a distraction from making a better society here and now. But I try to understand why others think it's so important. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 13,226 | 1.310811 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf4p3l | ivfnk1y | 1,667,832,901 | 1,667,840,514 | 68 | 97 | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 7,613 | 1.426471 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivfgxyt | 1,667,840,514 | 1,667,837,919 | 97 | 61 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 1 | 2,595 | 1.590164 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfb7ap | ivfnk1y | 1,667,835,618 | 1,667,840,514 | 59 | 97 | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 4,896 | 1.644068 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveu9yt | ivfnk1y | 1,667,828,142 | 1,667,840,514 | 50 | 97 | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 12,372 | 1.94 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivesaad | 1,667,840,514 | 1,667,827,120 | 97 | 50 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 13,394 | 1.94 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | ivfnk1y | 1,667,838,114 | 1,667,840,514 | 41 | 97 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 2,400 | 2.365854 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivfnk1y | 1,667,828,186 | 1,667,840,514 | 31 | 97 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 12,328 | 3.129032 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivfena9 | 1,667,840,514 | 1,667,837,000 | 97 | 29 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Neuroscience: There is no free will. | 1 | 3,514 | 3.344828 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivfnk1y | 1,667,838,362 | 1,667,840,514 | 36 | 97 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 2,152 | 2.694444 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfkric | ivfnk1y | 1,667,839,428 | 1,667,840,514 | 28 | 97 | We all talk a good game about null findings and replication, but it's thankless and minimally funded work. People kinda care, but only when somebody else does it. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 1,086 | 3.464286 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfl10y | ivfnk1y | 1,667,839,532 | 1,667,840,514 | 29 | 97 | Sometimes I think that philosophy should turn all of the classic stuff --- everything from Socrates and Plato to (oh) Nietzsche and Russell --- over to history departments. I think this would be better for philosophy in the long run, but mostly I don't think we can be trusted with it. | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | 0 | 982 | 3.344828 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivf10s1 | 1,667,840,514 | 1,667,831,300 | 97 | 26 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 9,214 | 3.730769 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivfd2kp | 1,667,840,514 | 1,667,836,370 | 97 | 24 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | 1 | 4,144 | 4.041667 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfnk1y | ivfg4n4 | 1,667,840,514 | 1,667,837,592 | 97 | 28 | I'm in Ecology; species are fake. To be more precise, the Linnaean taxonomic system is outdated, and the old species/genus/family/etc titles are at best guideposts along the twisted limbs on the tree of life. We can now see those limbs in greater detail via genetic testing, which has re-shaped taxonomic groups extensively and continues to do so. As much as we love putting things into neat little boxes, the reality is that some taxonomic groups end nicely in what we'd call a single species (humans, western red cedar), while others definitely don't (ash trees, oak trees, roses, bacteria, dogs/wolves, cows/yaks). Moreover, the adherence to species-level identification in field work and in scientific literature is actively detrimental to the accuracy of the claims made by that work. An example from my specific work in forest community analysis; if I have a group of field workers working to survey a given plot and it's full of oak trees, then I know that in reality I do not have good determination between different oak "species". Even if all my field hands are graduate-level people, the decision just comes down to subjective judgements eventually. We would need genetic testing to really make a determination, and that's completely unfeasible. So I think ecological literature should move more towards recognizing that the utility of the species level ID is often limited, and does not deserve its position as the ubiquitous tool that it currently occupies. But that's all way too long, so; species are fake 🤣 | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | 1 | 2,922 | 3.464286 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf1b60 | iveslwa | 1,667,831,428 | 1,667,827,288 | 84 | 74 | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | I'm in religion, technically in 'theology' not religious studies. I train ministers and chaplains. I'm pretty sure God doesn't exist. More agnostic than atheist, but even if God does exist, I don't think They matter that much. I think all the time angsting over who God is or what God wants is a distraction from making a better society here and now. But I try to understand why others think it's so important. | 1 | 4,140 | 1.135135 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveu9yt | ivf1b60 | 1,667,828,142 | 1,667,831,428 | 50 | 84 | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | 0 | 3,286 | 1.68 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf1b60 | ivesaad | 1,667,831,428 | 1,667,827,120 | 84 | 50 | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 4,308 | 1.68 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivf1b60 | 1,667,828,186 | 1,667,831,428 | 31 | 84 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | 0 | 3,242 | 2.709677 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf1b60 | ivf10s1 | 1,667,831,428 | 1,667,831,300 | 84 | 26 | Psychologist. All of the implicit bias stuff has low validity and reliability and has wasted millions of dollars. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 128 | 3.230769 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf59r0 | iveslwa | 1,667,833,145 | 1,667,827,288 | 78 | 74 | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | I'm in religion, technically in 'theology' not religious studies. I train ministers and chaplains. I'm pretty sure God doesn't exist. More agnostic than atheist, but even if God does exist, I don't think They matter that much. I think all the time angsting over who God is or what God wants is a distraction from making a better society here and now. But I try to understand why others think it's so important. | 1 | 5,857 | 1.054054 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf4p3l | ivf59r0 | 1,667,832,901 | 1,667,833,145 | 68 | 78 | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | 0 | 244 | 1.147059 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf59r0 | iveu9yt | 1,667,833,145 | 1,667,828,142 | 78 | 50 | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | 1 | 5,003 | 1.56 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf59r0 | ivesaad | 1,667,833,145 | 1,667,827,120 | 78 | 50 | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 6,025 | 1.56 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivf59r0 | 1,667,828,186 | 1,667,833,145 | 31 | 78 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | 0 | 4,959 | 2.516129 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf59r0 | ivf10s1 | 1,667,833,145 | 1,667,831,300 | 78 | 26 | Anthropology has the unique ability to counter the racist narratives and beliefs that its responsible for creating in the centuries past. The current emerging beliefs and endorsements in anti-anti-racist rhetoric, White supremacy, race patterned disparities, and the like have a special place in Anthropological education, and can be effectively combated by educating the public on the nature of the human condition and all that it does and doesn't entail. Instead, Anthropology suffers from an overrepresentation of White liberals who will continue to focus on non-White others, placing them on shelves to collect dust, and studying *anything other* than the ways that they themselves reproduce that culture of racial homogeneity within the academic field -- rendering Anthropology virtually useless and undermining any Anthro department's ability to secure funding, advance research in critical areas, or do anything of substantive importance beyond its basic requirements of studying what it means to be human on a biological and cultural level. What a waste. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 1,845 | 3 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivesaad | iveslwa | 1,667,827,120 | 1,667,827,288 | 50 | 74 | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | I'm in religion, technically in 'theology' not religious studies. I train ministers and chaplains. I'm pretty sure God doesn't exist. More agnostic than atheist, but even if God does exist, I don't think They matter that much. I think all the time angsting over who God is or what God wants is a distraction from making a better society here and now. But I try to understand why others think it's so important. | 0 | 168 | 1.48 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfgxyt | 1,667,848,821 | 1,667,837,919 | 68 | 61 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 1 | 10,902 | 1.114754 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfb7ap | 1,667,848,821 | 1,667,835,618 | 68 | 59 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | 1 | 13,203 | 1.152542 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveu9yt | ivg98ko | 1,667,828,142 | 1,667,848,821 | 50 | 68 | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 20,679 | 1.36 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivesaad | ivg98ko | 1,667,827,120 | 1,667,848,821 | 50 | 68 | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 21,701 | 1.36 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfyz8k | 1,667,848,821 | 1,667,844,890 | 68 | 46 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 1 | 3,931 | 1.478261 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfhfxd | 1,667,848,821 | 1,667,838,114 | 68 | 41 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | 1 | 10,707 | 1.658537 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | iveud6r | 1,667,848,821 | 1,667,828,186 | 68 | 31 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | 1 | 20,635 | 2.193548 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfena9 | 1,667,848,821 | 1,667,837,000 | 68 | 29 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Neuroscience: There is no free will. | 1 | 11,821 | 2.344828 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivg98ko | 1,667,838,362 | 1,667,848,821 | 36 | 68 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 10,459 | 1.888889 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfkric | ivg98ko | 1,667,839,428 | 1,667,848,821 | 28 | 68 | We all talk a good game about null findings and replication, but it's thankless and minimally funded work. People kinda care, but only when somebody else does it. | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 9,393 | 2.428571 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfl10y | ivg98ko | 1,667,839,532 | 1,667,848,821 | 29 | 68 | Sometimes I think that philosophy should turn all of the classic stuff --- everything from Socrates and Plato to (oh) Nietzsche and Russell --- over to history departments. I think this would be better for philosophy in the long run, but mostly I don't think we can be trusted with it. | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 9,289 | 2.344828 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivf10s1 | 1,667,848,821 | 1,667,831,300 | 68 | 26 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 17,521 | 2.615385 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivg6nir | 1,667,848,821 | 1,667,847,803 | 68 | 29 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Nuclear structure: literally stamp collecting. Who are these 100+ gamma-ray transitions at 0.001% intensity going to help? No one, that’s who. Pain in the butt to end up with a level scheme with 300+ transitions, just for it to be added to the database in 10 years. What’s the use? | 1 | 1,018 | 2.344828 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivfd2kp | 1,667,848,821 | 1,667,836,370 | 68 | 24 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | 1 | 12,451 | 2.833333 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfg4n4 | ivg98ko | 1,667,837,592 | 1,667,848,821 | 28 | 68 | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | 0 | 11,229 | 2.428571 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivg98ko | ivg59rp | 1,667,848,821 | 1,667,847,273 | 68 | 22 | Social work: That the answer to most of our research questions is this: people need more resources and structural systems/capitalism will never allow it. Individual interventions only go so far and will never fix the problems that throwing money at them will. All of our implications sections should just say: we need a more equitable society/eat the rich. But people won’t say it. | I don't work in this field anymore, but did grad school in cognitive neuroscience, primarily focusing on cortical development. Neuroscientists greatly exaggerate how authoritative MRI/fMRI research is at answering questions that are essentially philosophical questions. I'll go a step further and say that most neuroscientists use anatomical phenomena to support things about philosophy/cognition they they already believe. These studies tend to have small samples and feature observations like "subject did thing, this area of the brain lit up, that's the thing area". They also tend to be done by labs that have good anatomical and biological knowledge, but middling tech and statistics knowledge, so they're rife with misunderstandings about what results show. There are other examples of things like this that aren't related to fMRI research, for instance the classic study about electrical signals from the brain pre-empting conscious choice that has shown to have numerous errors, but nevertheless is constantly trotted out to prove that free will doesn't exist. I'm not religious or anything or have any skin in that game, but it annoys me to see how flippant scientists are about causality in this specific circumstance. Many are willing to take that extra step and just assume which way the arrow points because viewing brain activity as the result of a different phenomenon, or anything more complex than the summation of action potentials, is too woo for people. It's especially annoying to me because neuroscience does only the most rudimentary job of explaining consciousness, and only with strokes so broad that it almost becomes pointless to investigate in the first place. And on the topic of how tremendously complex the behavior in question is, views are something like "yeah, but you can't possibly explain every variation in human behavior, that's an unreasonably high bar to clear. Also though, we've cleared it." People want to have it both ways: acknowledge the complexity of the problem for the purposes of never needing to explain everything (or in some cases, anything), yet simultaneously simplifying the problem for the purposes of putting the issue on the shelf. Neuroscientists are more than happy to make vast, sweeping claims about philosophy, religion, free will, whatever, but as soon as you impose the high burden to settle those philosophical questions definitively, neuroscientists back down from that ledge and start getting sloppy about what's needed to settle the issue. It's one thing to say "based on which way the wind is blowing here, I know what I think is the most reasonable explanation, and that's good enough for me", but it's another to pretend like the data definitively proves "big question" things about the brain's relationship with consciousness/cognition. | 1 | 1,548 | 3.090909 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveu9yt | ivf4p3l | 1,667,828,142 | 1,667,832,901 | 50 | 68 | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | 0 | 4,759 | 1.36 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf4p3l | ivesaad | 1,667,832,901 | 1,667,827,120 | 68 | 50 | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 5,781 | 1.36 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf4p3l | iveud6r | 1,667,832,901 | 1,667,828,186 | 68 | 31 | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | 1 | 4,715 | 2.193548 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf4p3l | ivf10s1 | 1,667,832,901 | 1,667,831,300 | 68 | 26 | Most of these "unpopular opinions" are really quite popular throughout academia imo. We're clearly just united in our misery, irreproducibility, real-life applicability, and out-of-touch HR departments. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 1,601 | 2.615385 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfb7ap | ivfgxyt | 1,667,835,618 | 1,667,837,919 | 59 | 61 | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 0 | 2,301 | 1.033898 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfgxyt | iveu9yt | 1,667,837,919 | 1,667,828,142 | 61 | 50 | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | 1 | 9,777 | 1.22 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfgxyt | ivesaad | 1,667,837,919 | 1,667,827,120 | 61 | 50 | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 10,799 | 1.22 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivfgxyt | 1,667,828,186 | 1,667,837,919 | 31 | 61 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 0 | 9,733 | 1.967742 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfena9 | ivfgxyt | 1,667,837,000 | 1,667,837,919 | 29 | 61 | Neuroscience: There is no free will. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 0 | 919 | 2.103448 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf10s1 | ivfgxyt | 1,667,831,300 | 1,667,837,919 | 26 | 61 | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 0 | 6,619 | 2.346154 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfd2kp | ivfgxyt | 1,667,836,370 | 1,667,837,919 | 24 | 61 | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | 0 | 1,549 | 2.541667 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfgxyt | ivfg4n4 | 1,667,837,919 | 1,667,837,592 | 61 | 28 | I'm in literary studies. Unlike lots of my peers, I believe that literary critics' purpose is to understand and explain how literature *works*, and not to say whether it's *good* or *real* literature. | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | 1 | 327 | 2.178571 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveu9yt | ivfb7ap | 1,667,828,142 | 1,667,835,618 | 50 | 59 | The findings of most human-environment research point in directions that are economically inconvenient. Politicians are aware of what we're saying and what needs to be done but don't care because it would affect their fundraising and ability to be re-elected. | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | 0 | 7,476 | 1.18 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfb7ap | ivesaad | 1,667,835,618 | 1,667,827,120 | 59 | 50 | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | I'm in public health nutrition. Sugar-sweetened beverage (SSB) taxes are gross. People talk about how they are super effective as a public health intervention because "low socio-economic status populations bear the brunt of the burden of disease associated with SSB and they are most 'price sensitive' to the taxes and so they are most effective where they are needed most." I don't *deny* any of that's true, but hearing all the "equity is important, we must respect the individual!" public health practitioners basically say "This intervention works by leveraging the fact people are poor to make them behave in the way we think they need to," is *so* gross and is textbook paternalism. If asked about this aspect they'll say that the taxes can go back into public health interventions that will better the health of the community like that makes it less gross. It doesn't. Fund your interventions in ways that don't raise funds by burdening those already so burdened in the name of helping them. | 1 | 8,498 | 1.18 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivfb7ap | 1,667,828,186 | 1,667,835,618 | 31 | 59 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | 0 | 7,432 | 1.903226 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf10s1 | ivfb7ap | 1,667,831,300 | 1,667,835,618 | 26 | 59 | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | There is no theoretical justification for the existence of some political science subfields. In particular, the "American Politics" subfield shares pretty much all of its theory with the Comparative Politics subfield. The key difference between the two is that the former focuses on the politics of America and the latter is focused on the politics of every other country. I don't think it is unreasonable to have a regional specialization, but separating American politics into a separate subfield creates perverse incentives. For instance, to get an academic job you may be forced to clarify whether you are an "Americanist" or a "Comparativist" by either focusing your research on the United States in the former case or not researching it at all in the latter. | 0 | 4,318 | 2.269231 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfyz8k | ivfhfxd | 1,667,844,890 | 1,667,838,114 | 46 | 41 | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | 1 | 6,776 | 1.121951 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | iveud6r | ivfyz8k | 1,667,828,186 | 1,667,844,890 | 31 | 46 | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 0 | 16,704 | 1.483871 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfyz8k | ivfena9 | 1,667,844,890 | 1,667,837,000 | 46 | 29 | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | Neuroscience: There is no free will. | 1 | 7,890 | 1.586207 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivfyz8k | 1,667,838,362 | 1,667,844,890 | 36 | 46 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 0 | 6,528 | 1.277778 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfyz8k | ivfkric | 1,667,844,890 | 1,667,839,428 | 46 | 28 | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | We all talk a good game about null findings and replication, but it's thankless and minimally funded work. People kinda care, but only when somebody else does it. | 1 | 5,462 | 1.642857 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfyz8k | ivfl10y | 1,667,844,890 | 1,667,839,532 | 46 | 29 | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | Sometimes I think that philosophy should turn all of the classic stuff --- everything from Socrates and Plato to (oh) Nietzsche and Russell --- over to history departments. I think this would be better for philosophy in the long run, but mostly I don't think we can be trusted with it. | 1 | 5,358 | 1.586207 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfyz8k | ivf10s1 | 1,667,844,890 | 1,667,831,300 | 46 | 26 | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 13,590 | 1.769231 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfd2kp | ivfyz8k | 1,667,836,370 | 1,667,844,890 | 24 | 46 | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 0 | 8,520 | 1.916667 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfg4n4 | ivfyz8k | 1,667,837,592 | 1,667,844,890 | 28 | 46 | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | English literature. My unpopular opinion is that reviewers/editors need to give more attention to language in articles. I’m finding it harder to read/enjoy criticism due to so many buzzwords being thrown about and misused. Get to the point and be concise! Prove your intellect by directly presenting your ideas rather than relying on flowery language. | 0 | 7,298 | 1.642857 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | iveud6r | 1,667,838,114 | 1,667,828,186 | 41 | 31 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | 1 | 9,928 | 1.322581 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfena9 | ivfhfxd | 1,667,837,000 | 1,667,838,114 | 29 | 41 | Neuroscience: There is no free will. | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | 0 | 1,114 | 1.413793 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | ivf10s1 | 1,667,838,114 | 1,667,831,300 | 41 | 26 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 6,814 | 1.576923 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | ivfd2kp | 1,667,838,114 | 1,667,836,370 | 41 | 24 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | 1 | 1,744 | 1.708333 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfhfxd | ivfg4n4 | 1,667,838,114 | 1,667,837,592 | 41 | 28 | Computer Scientist here. My main research area is machine learning, and my sort of unpopular opinion is that deep learning isn't very interesting and is kind of a dead end. It's been the only thing getting any oxygen for about 15 years now, and it's still providing a lot of practical advancements, but all those advancements are being gained by just throwing money at it. To play at all in the space requires you to work for one of half a dozen companies that can afford to spend $20 million on a single training pass of one model. And we know that there are significant gaps that can't be closed without some fundamental shift in the approach, but because current deep learning models are miles better than the competitors, it's not feasible for most people to give up those gains to focus on something different that might have longer term payoffs. By sheer volume of research, that's an unpopular opinion, but within the field, I haven't really said much that would be objectionable to a lot of people either. For a truly unpopular opinion in the broader field of CS and software engineering, I think the state of software development is worse today than it was in the 1980s. Prior to that point, programming operated a bit more like an art form. It wasn't so much that people consciously thought of it that way. It was, I think, more a combination of a much smaller group of people doing it combined with a lack of conscious effort to frame the problem any other way. Good code was judged in similar ways as a good novel. A writer you love may have a great facility for poetic phrasing, and that's the thing you like. That's hard to quantify though, and I think code has a similar quality. As software engineering became more of a separate discipline that was focused on metrics like code reuse, it put objective-looking metrics in place of what had previously been a much more aesthetic process of judging code quality. And I'm not sure if I think the metrics were just bad or whether the failure was in not balancing optimizing for them against something else, but either way, programming got worse both in process and result. Give me a program written in the 80s with no thoughts at all toward code reuse, design patterns, dependency injection, unit test coverage, mocks, and all the other ostensibly great things about modern software development. If that code was written by a good programmer who had solid aesthetic judgements about how his code should be structured and written, it will be easier to work on, more bug free, faster, etc., than almost any software written today on the best engineering principles we've been able to come up with in the last 30 years. And if it were written by a bad programmer, it would be bad, but it would be more obviously bad than the average "modern" codebase. | History: there's a *long*-standing insistence in the field that we need to be "understandable" to the broad public. Overly-academic writing is thus "bad", and we should all be striving to popularize our work and our field as much as we can through social media, popular history writing, etc. I'm sorry, but History has long been far too specialized for that to be truly successful. We are building on generations upon generations of scholarship, while the public just occasionally watches a Ken Burns documentary at best... No one without a PhD in History is going to understand *anything* about true cutting-edge historical scholarship. No amount of social media posting, no amount of dumbing it down, could ever begin to distill the *thousands* of books in my subfield that I needed to fully understand just to *start* my own research. Sure, Historians can offer "corrections" to the most gross misinterpretations... EDIT: and yes, we can teach intro and advanced classes, and publish popularizations of our material. And that's good! But, *like every other professional field, from astrophysics to microbiology*, our serious work--our scholarship--is too specialized and technically-sophisticated to be understood by the uneducated masses or even by the chattering classes. (or even, truth be told, by other fields. The number of times I've encountered a mathematician or physicist who thinks that they know *anything* about my field--and eventry to lecture *me* on it!--is just way too many times.) | 1 | 522 | 1.464286 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | iveud6r | 1,667,838,362 | 1,667,828,186 | 36 | 31 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | Lots in education, somehow, still think that schooling is primarily a psychological process in which social forces don't matter. Similarly, Theories and approaches, especially in contemporary neoliberal america, get stripped of their original social and political commitments. I saw a syllabus for a course on Critical Pedagogy and it didnt have a single reading by Freire or any other progressive. Critical Literacy gets transformed into "Critical Thinking" | 1 | 10,176 | 1.16129 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivfena9 | 1,667,838,362 | 1,667,837,000 | 36 | 29 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | Neuroscience: There is no free will. | 1 | 1,362 | 1.241379 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivf10s1 | ivfena9 | 1,667,831,300 | 1,667,837,000 | 26 | 29 | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | Neuroscience: There is no free will. | 0 | 5,700 | 1.115385 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfena9 | ivfd2kp | 1,667,837,000 | 1,667,836,370 | 29 | 24 | Neuroscience: There is no free will. | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | 1 | 630 | 1.208333 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfi2ex | ivf10s1 | 1,667,838,362 | 1,667,831,300 | 36 | 26 | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | Computational chemistry, molecular dynamics: You can run a bunch of methods with a bunch of target properties on a single dataset and find "something". The problem is that often the size of the dataset should increase the more properties you study. In other words, you are more likely to see a fluke with a medium dataset if you use a lot of approaches. To counter this issue (ie hacking) in clinical trials Registered Reports have been introduced. | 1 | 7,062 | 1.384615 |
yolicw | askacademia_train | 0.97 | What's your unpopular opinion about your field? Title. | ivfd2kp | ivfi2ex | 1,667,836,370 | 1,667,838,362 | 24 | 36 | Organization Studies/Science (yes, it matters, and i used to be active at the intersection). I'm no longer active as a researcher, luckily. Not only is most research utterly irrelevant for people managing organizations, I've come to find it increasingly less important what people managing organization need or want. Fuck 'm. | That studies of intelligence (as in the activities of intelligence agencies, not human or animal intellect) should be properly viewed a sub-field of history because its occasional broader theoretical claims tend to be deeply suspect or so obvious that they offer very little insight. I’m ok with this, personally, but lots within the discipline see it otherwise. | 0 | 1,992 | 1.5 |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.