timkam

I agree. To me, the post reads a bit like this: "Back then, we did not have as much inequality, but the super-rich did not deserve their wealth. Now, we have more inequality, but the super-rich [like PG himself] really deserve it, because they are innovators." The post does not cover the arguably more important questions like "Are the working and middle classes better off?", "Was there less innovation back then or were lesser paid innovators just as innovative?", and "Can we have as much or even more innovation with less inequality?".

Do you know whether Cree# is open source? Sorry if I missed this when reading the text too quickly, but I struggle to find more documentation or a code repository. I know some folks at my university who are working with indigenous languages and cultural heritage, this could be interesting for them.

I think the author is tearing down the brand rather than the person. And I think it's good. For sure, Hawking's individual productivity was limited in later years, as was his control of what would be associated with this name. For sure, Cambridge tried to profit from Hawking's name in ways that are only remotely related to his research. For sure, other, similarly talented researchers contributed to many of Hawking's achievements.

IMHO, it is important that somebody discusses these issues, because we should not glorify the intellect of individuals to such an extreme extent. Woit might hold some research-/community-related grudges, but I don't think this makes his post less relevant.

This is certainly important from a business perspective. For-profit publishing is a dead end. More and more venues are community-owned, and only the highest profile publications actually need the degree of professionalization that requires a large publishing house. By now, it is clear that the publishers cannot win this fight. And we see this in the concessions publishers make. IMHO, publishers can either milk the cow as long as it's possible and then die, or transform their business. My university just announced an Open Access agreement with Elsevier, and to me this looks like a confession on Elsevier's side that they will have to go OA in one way or another, or they'll lose their journals one by one. In my community, Elsevier has already lost journals to community-owned replacements.

Sure, but I thought was that building an ML model that considers the questionnaire results (or existing similar findings) would be a very interesting next step.

Now, what would be interesting is to build a model that accounts for this "knowledge" and see if it can beat the out-of-the-box classifier :-) I'd assume it can, and the question is: how far can a bit of manual modeling bring us?

If the value of gold increased as rapidly as the value of bitcoin, and the value of all kind of "I pull a fictitious metal name out of the air and make it tradable" materials increased at a striking pace as well, we'd also say it reeks of scam, right?

The author is not precise here, but I agree that in the crypto-bro, get-rich-quick world, the meaning of "decentralized" and "blockchain" certainly can be stretched in all kinds of ways. And I think this relates to the problem around cryptocurrencies: its communities are an unhealthy mix of a lot of people who want to make a quick bug, and a few people who are seriously working on the foundations of the technology in a meaningful manner.

I don't think it's about whether Gebru "deserves" the credit, it's just a bad comparison on so many levels. Roughly like: "the transition from Ronaldo to Magnus Carlsen may be seen in the same way as we see the transition from Pelé to Lebron James". I also think it would be fair to speak in terms of communities and not of individuals, at least in modern science.

I still think this post is insightful, because it shows a subjective and presumably rather honest perspective on reviewing CVs. Sure, it's no big deal that a recent graduate points to an unmaintained GitHub profile with only assignment code on it. But these applicants make whoever is considering hiring them look at something that's both irrelevant and boring. I wouldn't blame them too much, though. They've probably heard somewhere that you should have code on GitHub and link to your GitHub account in your CV, and they lack the experience to take this advise with a grain of salt.

I largely agree with this comment, but this just means the underlying issues are more subtle. For example:

* I typically get thoughtful reviews, and in particular critical reviews are often thoughtful. Still, I know that some people assert influence on the outcome of the reviewing process of their own papers. Also, there's a fine line: for example, many journals allow you to recommend reviewers for your papers. If you're well-connected, you can recommend your best friends and your editor friend follows your advice. Is this misconduct? I'd say it depends. After some decades in a math-heavy community, most people who are experienced and competent in the context of a highly specialized sub-field will be your "friends".

* Journals that are outright predatory are, of course, a no-go. Still, there are edge cases like some MDPI journals; sometimes there are special issues from credible, relatively junior people in my community. Should we support them or not? Also, my funding provider wants me to publish in high-profile, "big publisher" journals and pays the open access fee; these journals typically rank much higher than community-owned/not-for-profit venues. On the long run, it's better to fully move to community-owned venues only, but as a junior researcher I feel I cannot afford it at the moment.

Personally, for me the problem does not come from outright and obvious ethics violations; sure, I know of some individuals who I consider somewhat problematic, but I don't think these people control the community. For me the problem is to navigate all the grey areas, knowing that the system is largely ethics-agnostic.

I think this topic is more about what should be the focus of discussion in the AI community. A key question is: to what extent do AI researchers need to consider ethics and how do they need to consider it? These are hard questions. People in the broader AI community are typically no ethics experts, so the discussions can easily become uninformed and unpleasant, and some people on the extremes of both sides will try to use it to strengthen their profile (or perhaps, some of them simply can't hold their horses). Funding sources are a different topic, and many funding providers fund both AI intersectionality and research towards more effective autonomous weapons systems...

I really don't get it. Nobody at an AI conference is alarmed at what they see in terms of AI research progress. Most well-educated people are alarmed about societal issues that _are also relevant to the application of automated decision-making and machine learning_. It's an applied research opportunity to investigate the issues (and as such, it is hyped because it is easy to "sell" to get funding); but it's not that AI research brings "bias" that didn't exist in society before. It's the same as for a rule-based system, or let's say, even a programming language: for sure, it can automate and hence exacerbate "unfair" decisions, but the problem is not in the specification, but in the specifying person/human. However, having a clear specification makes it easier to test/audit/verify and hence we (as a society) should do so. We don't need to burn down academic prototypes just because they make some naive assumptions about the application domain, though.

Private schools have more leverage on whom they will accept as pupils and hence can target an "easier to teach" part of the overall pupil population.

To me the Swedish system is the worst of two worlds: in Germany, there are good public schools for well-performing students; the problem is that affluent parents typically make sure their children are classified as "well-performing" (roughly speaking; there is a lot of wiggle room). This is somewhat unfair, but facilitates to some extent the development of pupils with reasonably high potential. In other Scandinavian countries, the state provides enough resources to have small class sizes in which pupils of all abilities get educated together. This is fair but can hamper the development of high-potential children (teachers still have limited time and attention). In Sweden, the distribution of resources as criticized by the article facilitates that pupils in need don't receive special support and high-potential children get their education in for-profit institutions for pseudo-elites where a teacher is discouraged to give tough feedback because it can hurt the organization's bottom line.

Then don't take offense by the discussion here, because it's mostly on some "meta" aspects of science communication, and you are probably not responsible for any of the aspects that have been critizised.

Regarding the research itself, I am not an expert, but I am curious to learn how this line of research (automated conjecture generation) intersects with proof automation/proof assistants, and in particular with the work that the Lean community is doing (creating an "executable" collection of mathematical knowledge). Perhaps there are some works you can point to.

I think the vast majority of criticism here does not target the research per se, but rather the way the results are "hyped" and presented as a massive break-through. I agree with this criticism, and also think that the two positive Nature reviews seem rather shallow, at least from a non-expert's perspective (this is not your fault, of course). When it comes to long term impact, I'd find it interesting to discuss how your work can (ideally) interact with proof assistants like Lean. Also, the work around Lean is a good example of a "hyped" topic that is presented by its contributors with caution and modesty.

I agree. Without being an expert in the field, the 2nd and 3rd reviews "smell funny"; they clearly lack depth, are very enthusiastic, and don't seem to make a good and comprehensive point as to _why_ this paper is supposed to be as great as they claim it is. In a strong community, any editor should consider these reviews disappointing, and any author should at least have mixed feelings about them.

> Even more with some irresponsible newspapers blaming the whole thing on GamerGate and Alt-Right, effectively pushing away people that previously could be on their side.

I didn't read any such claim in my German and Scandinavian news diet, is this a US-only thing? Here, the consensus seems to be that what happens is a somewhat well-deserved embarrassment.

I agree with this concern. They collect a lot of data, sure, but they are also the only one who thinks that video is enough. It's a bold bet; intuitively, one would expect that video-only is good enough for gimmicks and fares worse when it comes to covering the millions of edge cases that still need to be addressed.

I think this is very interesting but not too surprising. Until a couple of hundred years, many (if not most) educated people where closely associated with some form of organized religion, and the study of human behavior from both descriptive and prescriptive perspectives was an important line of work of religious scholars. Also, I'm wondering if one can track a clear line from the principles of (micro)economic rationality, i.e. clear and consistent preferences, back to religious principles like "if in doubt, remain consistent" (colloquialized) as devised by Talmudic scholars in the 19th century, see e.g. Shev Shema'tata: https://en.wikipedia.org/wiki/Shev_Shema%27tata.

By adding an author profile and a list of other Singularity Hub articles by the author, they make it look like the author is part of the Singularity Hub community, but the two articles that he has written are both from The Conversation.

Thank you for clarifying, then the main issue comes with global scale (a point that one of the other comments here makes).

I must say the Singularity Hub web presence looks snake-oily to me. Why do they call it a "university"? Why do they call some of the members "faculty"? Why do they claim Ray Kurzweil has "a thirty-year track record of accurate predictions"? (https://singularityhub.com/experts/)

Any idea why they brand it like this? Who is supposed to take this seriously?

I'm EU-based and here you're not allowed to claim that your organization is a university if it is not recognized by the state as such (roughly speaking).

Self-citations are often hard to avoid (in particular in more formal academic writing, in which you want to establish the delta of your current work wrt. your previous works). It is striking, however, that the author exclusively cites himself. This is kind of weird, because it makes it look like the line of work is not related to anything anybody else is doing (lacks relevance).

Sure, but my question is about the difference between Norway and Sweden. Sweden has (merely) twice the population size of Norway, is culturally similar, highly developed, has a similar political system etc, but still seems to fair far worse.

What I find striking are the reported recidivism rates of Sweden and Norway: 43% vs. 20% after 2 years. Sure, Norway is richer and socially/ethnically more homogeneous, but can this be the whole explanation for this extreme difference, considering the two countries are otherwise very similar?

I largely agree, but we do not write all these papers "so tax payers can sleep". We write these papers because the administration incentivizes us to get the highest profile publications with the lowest effort possible. In relatively creative fields (i.e., much of computer science), this can even mean to make friends in the community and co-author papers with them without being the one who does any of the hard work. (In other sciences, it means being the one who pulls in the money, as far as I understand.)

A related problem I see is that we treat all papers the same: position (opinion) papers, reviews & surveys, and "hard" results (relevant proofs, strong empirical results, software artifacts that are deployed at scale in practice etc.). As the blog post somewhat suggests, the administration should primarily ask us to summarize these results when assessing whether we are worthy of funding and faculty positions; sure, it takes more effort, but I think it is possible.

One can perhaps say: "A neural net is a collection of machine learning approaches that are somewhat inspired by the human brain. However, neural nets have little to do with how the brain actually works." I think this statement (or a similar, more accurate version) is actually useful, because the name "neural net" is somewhat deceiving.

From a computer science perspective, this is an interesting piece. But from a practical perspective, I doubt that the Dijkstra of 1988 has a good understanding of the sociotechnical reality of the present day software industry. I comment (quite obviously) not to disparage him and his outstanding achievements, but rather to highlight that we cannot expect that Dijkstra could predict the future, in particular in a realm that somewhat exceeds the domain of his expertise. For example, I think by now we have a good understanding of the relevance and importance of software maintenance. Sure, software is not "subject to wear and tear", but the world around software evolves and a program that did a job well 10 years ago might not to the same job well today (think about security). We all understand this difference; i.e., I don't think his nit-picky attacks on software engineering metaphors are particularly useful from today's perspective.

I agree, and I hope to see more fully open review approaches in the future, where everyone who is interested can see the reviews and the reviewers' names. The Journal of Open Source Software has this: https://joss.theoj.org/. It's certainly not a high-profile journal, yet, but it starts getting contributions from strong researchers. Considering that in small communities, it is not uncommon that even the author can infer who the reviewer is, anyway, I think this is the way to go.