> performance doesn't matter for the staff to be paid.
That's not the case for the people who are in the key phase of establishing themselves as academics and who are generally not on permanent contracts. And for tenured staff, performance decides access to resources that may be as mundane as one's own research time.
Conversely, the problem is a mix of extremely shallow performance metrics and a paper-pusher system in which most people get bogged down by bureaucracy by the time they get tenure (or even earlier). This means that very few people have the peace of mind that allows them to focus on high-risk fundamental research. In European AI and CS, I would claim that the ones who manage to maintain focus essentially do this in their free time and get little reward for this; the alternative, i.e., building a strong franchise, is more rewarding in terms of performance metrics.
Note that "false positives" may appear in the Google Scholar results for such queries. For example, a high-profile AI researcher shared a screenshot of https://scholar.google.com/scholar?as_sdt=0%2C5&as_ylo=2023&... (on X/Twitter: https://twitter.com/MelMitchell1/status/1768422636944499133).
The second entry is an ironically written abstract of a talk and the author clearly (and consciously) plays with the cliché. Obviously, most of the fans of the person who shared the screenshot to thousands of followers will not notice this.
Is the adjacency in Scholar limited to intra-sentence fragments? It'd be problematic if not, certainly. Here is another example of the sibboleth: when we use "certainly" here. Is this a query hit?
The article somewhat suggests that the guy used the airport's wifi. In this case and assuming that the wifi network supports government surveillance by design, end-to-end encryption does not help.
That is simply not correct. With properly implemented E2EE, you can communicate confidently over a completely insecure channel. You could post the entire data stream publicly on the internet with no loss of privacy.
I would still assume that with apps like Snapchat, there are inherent problems on meta-level (from an E2EE view) because for the users, Snapchat is essentially a trusted third party, providing governance/management features. But it can well be that Snapchat does not provide E2EE; I cannot see any remarks about it on their webpage.
"WhatsApp _considers_ chats with businesses that use the WhatsApp Business app or manage and store customer messages themselves to be end-to-end encrypted."
And later, they write that in many cases, Meta can actually read the messages.
What I disagree with in the article is the _I never make mistakes_ attitude; it could be worse, but I still think it's good to discuss. The author writes that because they are "serious", they can essentially rebut all criticism "easily". We are all human and even excellent scientists make honest mistakes. Strong theorists can sometimes make "hard" math mistakes. In the best cases, peer review gives us some assurance that we at least did not make obvious mistakes that can be relatively easily spotted by other specialized researchers.
I think the _never make mistakes_ attitude is dangerous, because it means that researchers need to be very cautious when admitting honest mistakes and their own intellectual fallibility in order to not lose face.
If so, let's start with the former (god) and point us not only to the proof, but also to the theorem and the underlying definitions. There are many logicians with a scientific interest in god, religion, and related concepts. Why not be open minded, after all? But we should expect more rigor in this case.
A well-intended interpretation is: the parent comment advocates for a society where there is substantial language diversity even within smaller geographic regions.
This makes sense from many perspectives, because forcing people to speak the same language no matter the context is not only hostile to immigrants but also to native minorities. E.g., in the North of Scandinavia (where I reside), many people don't like that the language that has to be spoken is traditionally (less now than historically) dictated by the South.
> This makes sense from many perspectives, because forcing people to speak the same language no matter the context is not only hostile to immigrants but also to native minorities.
in terms of non-native languages, practically this sort of policy quickly becomes a nightmare in terms of ensuring equitable access to public services
either these people become disadvantaged in terms of dealing with the state, or small fortunes have to be spent hiring substantial amounts of interpreters for every possible situation (which is a significant cost for e.g. elementary schools)
Fully agree. Central functions of these types do not scale. Even with more mundane objectives, like operational excellence, organizations have learned that centralization leads to ivory tower nothing-burgers. Most of the resources should go to where the actual work gets done, as little as possible should be managed centrally (perhaps a few ops and thought leadership fluff folks...).
And decentralized functions tend to be wildly inconsistent across teams, with info sec being a particular disaster where I've seen that tried. Neither model is perfect.
Sure, but we are talking about research teams here, not about an ops or compliance team. Central research tends to be detached from the business units but does not provide any of the 'consistency' benefits. Central research makes sense if the objectives are outward-facing, not if one wants to have an effect on what happens in the software-building units. So I'd say that ideally/hopefully, the people of the RAI team will now be much closer to Meta's engineering reality.
It works for things you can automate. For example, at Microsoft they have some kind of dependency bot such as when you have newtonsoft installed but have version < 13.0.1 and don't upgrade within such and such time frame, your M1 gets dinged. This is a very simple fix that takes like five minutes of work if that.
But I don't know if things are straight forward with machine learning.
If the recommendations are blanket,
And there is a way to automate checks,
It could work.
Main thing is there should be trust between teams.
This can't be an adversarial power play.
Thanks. Looked into it, but they don't seem to be hiring MLEs/Researchers.
Also it seems to be a very generic, "we give you the AI" business without a clear product or service.
The general concept of luxury beliefs is very interesting, but it feels the author gives it an ideological spin:
1. The author claims that when luxury beliefs 'trickle down' to the working class, they are damaging, which is supposedly not the case for luxury goods. Still, in some not-so-privileged parts of society, signaling status with luxury goods consumes the larger parts of people's budget. Think of someone living in a cheap apartment while driving a (leased) luxury car. This is certainly more damaging than believing the police should be de-funded (which does not have any consequence whatsoever on the micro-level, stupid as it may be).
2. The author claims that advocating a disciplined work ethic while attributing one's own success to luck is inconsistent (and hence bigoted). Still, it _is_ true that success is mostly luck, due to opportunities we cannot control. This does not mean that one should not work hard. It's not that affluent people say: "I am rich because of luck, so you should not even try." The realistic take is that success is mostly due to luck but hard work is one of the few ways to control success _to at least some extent_. Also, in many societies hard work is seen as a virtue in itself, not matter whether it pays off or not. So I would claim that "hard work pays off" is a luxury belief for people who are already super privileged, have strong networks, et cetera. When you are born into the elite, you can afford claiming that marrying into the elite is inferior to working your way up...
This paper presents formal results, apparently. Also, in the case of formal results, peer review by experts in the exact sub-field makes sense/increases trust, sure, but why not share on archive instead of waiting for a year (or however long the process takes in the particular instance)?
That's not the case for the people who are in the key phase of establishing themselves as academics and who are generally not on permanent contracts. And for tenured staff, performance decides access to resources that may be as mundane as one's own research time. Conversely, the problem is a mix of extremely shallow performance metrics and a paper-pusher system in which most people get bogged down by bureaucracy by the time they get tenure (or even earlier). This means that very few people have the peace of mind that allows them to focus on high-risk fundamental research. In European AI and CS, I would claim that the ones who manage to maintain focus essentially do this in their free time and get little reward for this; the alternative, i.e., building a strong franchise, is more rewarding in terms of performance metrics.