Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Great AI Weirding

deliprao.substack.com

Post reply on HN
Print view View original

41–50 of 51 posts

Re: The Great AI Weirding

#41

Post by thrwayaistartup »

godelski wrote: post #19
thrwayaistartup wrote: post #15

Earlier quoted context omitted.

Or just leave academia. In the US at least, the job is like 80% government contracting and 20% teaching. Teaching is great, so there's that. But literally every company will let your ad junct, and Professor of Practice usually pays more than 20% of a faculty salary. You can supervise PhD students as interns or by taking a courtesy affiliation (and often even have more impact on those students than their overworked an…

> Or just leave academia The problem is that AI is weird not because of academia. In fact, right not it has been captured by industry and it is why we've severely slowed down in progress[0]. Most people in the space now are working in industry labs. Frankly, you can do more, you get paid A LOT more (2-3x) and you have less bureaucratic bullshit. But I think you're keenly aware of this industry capture as you're menti…

Only a relatively tiny sliver of PhDs doing top-tier ML research are in groups that care about publishing at corps the care about publishing in academic conferences.

Re: The Great AI Weirding

#42

Post by photochemsyn »

So, I guess we're all going to end up as rodents in a Skinner Rat Box being subjected to AI-managed operant conditioning in order to further our career goals? Surely this will lead to the creation of a Brave New World, free of all conflict and suffering - but only if we can ensure that AI is safe, properly aligned and guided by today's government leaders and corporate executives, who have no desires other than to benefit humanity, or so they assure us.

It's funny how so many designers of utopian paradises ended up creating dystopian hellholes, historically speaking, isn't it?

Re: The Great AI Weirding

#43

Post by hellisothers »

jamilton wrote: post #36
coffeecat wrote: post #33

Earlier quoted context omitted.

Therein lies the problem. Most people are apathetic and incompetent. Winning rat race points requires some amount of competence and "oomph", so rat race points will always correlate positively with applicant quality. Most rat-race-losers, unfortunately, aren't Socrates.

I wonder what other tests of being non-apathetic and at least somewhat competent would work as well or better than the typical rat race.

Anything that signals grit

Re: The Great AI Weirding

#44

Post by keiferski »

I think the logical and inevitable conclusion of this quantification is simply throwing out quantification as useful, especially when it comes to things like elite hiring decisions. What will replace it? Charisma, personality, and other attributes that simply can’t be quantified.

It’s easy to forget that the drive to make highly-quantified decisions is largely a recent phenomenon, with in-person charisma having a much longer history. The recent widespread dominance of online video (compared to text) is really just more of a return to this kind of charisma after a long period of textual dominance.

I think the future is dominated by people that understand how to use video (and way down the line, 3D presence tools), not those that are good at optimizing AI tools.

One example of this, I think, is how video searches on TikTok/YouTube seem to be replacing Google searches for younger people. The searcher of 2030 isn’t going to read a perfectly individualized AI-created blog post, they’re going to watch a video by someone they trust.

TLDR: widespread video will herald a return to charismatic authority, displacing quantification systems of authority.

Re: The Great AI Weirding

#45

Post by godelski »

thrwayaistartup wrote: post #40
godelski wrote: post #20

Earlier quoted context omitted.

Honestly, is there a big difference anymore? The vast majority of papers I read are either by industry directly or have industry as a partner (as an author, not just acknowledgements). There are of course some, and even plenty of examples, but it does seem industry partners is almost necessary these days. I'm not convinced that level of interaction is healthy, for either parties.

> Honestly, is there a big difference anymore? Only a very small subset of industry cares about academic publishing, and even within that subset it's only a fraction of groups at a fraction of corps that consider publishing a primary or even secondary objective. The groups that do care about those things can be good gigs, but are generally not the place in the company you want to be anyways, unless you can get in and…

You're measuring the wrong direction. Don't measure what percentage of industry publishes with academics. Instead, measure what percent of academics __in ML__ publish with industry. This direction because one is much larger than the other. Second, I mean... I am a researcher... and I'm talking about the environment I'm working in. It sounds like you're outside this environment trying to educate me on it. Am I misunderstanding here?

> can do something that actually impacts the business -- that is actually useful to other humans

Do not confuse these two. That's incredibly naive.

Re: The Great AI Weirding

#46

Post by bradleyjg »

gyomu wrote: post #11
bradleyjg wrote: post #4

We, collectively as a community, are forced to play these stupid RL career games because if you refuse, you become illegible and, consequently, invisible to sources of physical, emotional, and intellectual sustenance. It’s like we are all trapped in vicious cycles of RL career games while hoovering up others in these cycles. Once a critical threshold of people start playing these RL career games, these terrible metri…

I did that with a few interns/junior people and they turned out to be disappointing rather than predictably satisfying. I still agree with you, because I want to believe it pays off in the long run. But it definitely comes at a real short term cost.

I expect, and have a few data points to suggest, that the average is lower but the variance is higher.

More importantly you are helping to fight an anti-human machine that sacrifices endless years of (young! healthy!) life to pointless grinding.

Re: The Great AI Weirding

#47

Post by poulpy123 »

mpoteat wrote: post #3

The key, from my perspective, is to focus on real work which produces real value for other human beings. In an academic context I admit that it can be difficult to attach this value to your work. As well, in this article, a person bemoans the opportunities their parents gave them; piano lessons, math competitions, etc. Even though they have clearly benefited from these advantages, it's unclear to me if the person ack…

The author is not bemoaning the opportunities he had but the fact these became requirements

Re: The Great AI Weirding

#48

Post by aleph_minus_one »

godelski wrote: post #37
aleph_minus_one wrote: post #28

Earlier quoted context omitted.

> Honestly, is there a big difference anymore? The vast majority of papers I read are either by industry directly or have industry as a partner (as an author, not just acknowledgements). Read more pure math papers, then you will see the difference. :-)

I thought we were talking ML here. I mean you're not wrong (I do do this) but context. But in ML, well... I mean even Max Welling is connected with Microsoft.

> I thought we were talking ML here.

This is no contradiction: there exist quite pure math papers whose content is very relevant for the mathematics behand ML algorithms. :-)

I do have the impression that the kind of research in ML that is not strongly associated with the recent "machine-learning industrial complex" by now tends to become published in another subject area.

Re: The Great AI Weirding

#49

Post by feoren »

thrwayaistartup wrote: post #7

The open secret is that top-quartile R1 CS faculty positions aren't coveted anymore and don't attract the best like they used to. The choice is now between increasingly tenuous/meaningless tenure after 5-10 years and a $500K/year lower bound for 10-12 years. That choice is... not a hard choice for anyone who values intellectual freedom. And the right answer sure as shit isn't the faculty position. A good 50% of those…

> a $500K/year lower bound for 10-12 years

I don't care if you are talking about top talent here; that is an insane thing to say. As a lower bound? What percentage of software engineers / AI practitioners / data scientists are making $500k/year? 0.1%?

Re: The Great AI Weirding

#50

Post by godelski »

aleph_minus_one wrote: post #48
godelski wrote: post #37

Earlier quoted context omitted.

I thought we were talking ML here. I mean you're not wrong (I do do this) but context. But in ML, well... I mean even Max Welling is connected with Microsoft.

> I thought we were talking ML here. This is no contradiction: there exist quite pure math papers whose content is very relevant for the mathematics behand ML algorithms. :-) I do have the impression that the kind of research in ML that is not strongly associated with the recent "machine-learning industrial complex" by now tends to become published in another subject area.

Sure, I agree with you. I just wouldn't refer to that work as pure math. And let's be real, most people are not working on the theoretical side of ML. Realistically people are anti theory in the ML space and it's really weird to me because it's a self fulfilling prophecy and the complaints are "it's not very good because not a lot of community effort hasn't been put in so let's not waste our time"
Post reply on HN