T

Tsunayoshi

698 karmaJoined

Comments
87

Great post, we need more summaries of disagreeing view points! 


Having said that, here are a few replies: 
 

I think this makes me very concern of a strong ideological and philosophical bubble in the Bay regarding these core questions of AI

I am only slightly acquainted with Bay area AI safety discourse, but my impression is indeed that people lack familiarity with some of the empirically true and surprising points made by skeptics e.g. Yann LeCun(LLMs DO lack common sense and robustness), and that is bad. Nevertheless, I do not think you are outright banished if you express such a viewpoint. IIRC Yudkowsky himself asserted in the past that LLMs are not sufficient for AGI (he made a point about being surprised at GPT-4 abilities on the Lex Fridman podcast). I would not put too much stock into LW upvotes as a measure of AIS researchers POV, as most LW users are engaging with AIS as a hobby and consequently do not have a very sophisticated understanding of the current pitfalls in LLMs.
 


On priors, it seems odd to place very high credence in results on exactly one benchmark. The fate of most "fundamentally difficult for LLMs, this time we mean it" benchmarks has usually been that next gen LLMs perform substantially better at them, which is also a point "Situational Awareness" makes. (e.g. Winograd schemas, GPQA). Focusing on the ARC challenge now and declaring it the actual true test of intelligence is a little bit survivorship bias.   

Scale Maximalists, both within the EA community and without, would stand to lose a lot of Bayes points/social status/right to be deferred to

Acknowledging that status games are bad in general, I do think that it is valid to point out that historically speaking the "Scale is almost all you need" worldview has so far been much more predictive of the performances that we do see with large models. The fact that this has been taken seriously by the AIS-community/Scott/Open Phil (I think) well before GPT-3 came out, whereas mainstream academic research thought of them as fun toys of little practical significance is a substantial win.

Even under uncertainty about whether the scaling hypothesis turns out to be essentially correct, it makes a lot of sense to focus on the possibility that it is indeed correct and plan/work accordingly. If it is not correct, we only have the opportunity cost of what else we could have done with our time and money. If it is correct, well.. you know the scenarios. 

 
 

In general, there is no reason to expect the Atlas' founders to spend money needlessly. Nobody is suspecting that they are spending it on themselves (excepting the alleged expensive table), and just like enterprises I expect them to be at least trying to use their resources in the most efficient way possible.

You raise imho valid arguments. To address some of your points:

I guess the Atlas Foundation is going off a model where impact is heavy tailed, in which it makes sense to spend what seems disproportionate resources on attracting the most talented. In such a model, attracting a fellow from the 99th "potential impact" percentile rather than 10 fellows from the 95th percentile would still be worth spending some marginal 45k for, even though it sounds excessive.

"From friends who are Atlas Fellows, they said many Atlas Fellows do not require the scholarship as their parents earn a lot and can already pay for college." If true, this is evidence in favor of offering them such a ludicrous amount of money. They do not really need the money, so the marginal value is reduced and you need to offer more money to entice such potential students (or think of other benefits). And an unfortunate fact of life seems to be that a person's financial earnings are highly correlated with those of their parents. Taking earnings as a proxy for potential impact means that a program like the Atlas Fellowship should also consider privileged students as people worth attracting.

And maybe that's just me, but some of the phrasing comes off as somewhat combative (on the other hand I am aware that many people here think we should state our opinions more directly). As an example, the question in the title: "why do high schoolers need $50k each?" is not really truthful and sounds rhetorical, because nobody has claimed that the applicants need that money, just like high frequency traders do not need high compensation but still firms pay that amount to hire them.

I would usually not go around tone-policing, but I think it would be beneficial in controversial times like this to remember that as **a community we wanted to move away **from evaluating charitable initiatives based on how they sound and instead evaluate them on their results. In that vein, I do not think that it is helpful to quote rumoured single sentences by founders without any context ("not believing in budgets") and without actually engaging with that sentence. The founders do not owe us accountability of private sentences that they might have uttered at some point.

Hits based giving means that Open Phil should not police the furniture of their grantees, and I am also unsure whether the way they manage inventory is indeed of public interest, as they are not soliciting donations from the public at the moment.

Just wanted to point out that Peter and you seem to mention two different classes of behaviors. While the behaviors you mention certainly create a more unwelcoming environment to women and shouldn't be welcome in EA environmens, I don't think they would meet the (legal ?) definition of sexual harassment and may not be the types of actions Peter had in mind.

IANAL, so I will just quote another government website, but I would be very surprised if accusation 1 holds any water. This was not difficult to find at all; also it seems a bit odd to first admit to not understanding what "legal authority" means but bringing the accusation forth anyway.

" A charity can pay a trustee for the supply of any goods or services over and above normal trustee duties. The decision to do this must be made by those trustees who will not benefit. [...].

Examples of goods or services that may be provided by a trustee in return for payment under the power in the Charities Act include:

  • the delivery of a lecture
  • a piece of research work...
  • the occasional use of a trustee’s premises or facilities

"
https://www.gov.uk/government/publications/trustee-expenses-and-payments-cc11/trustee-expenses-and-payments#s5-3

it would be trivial for one of the staff to say so.

It seems unreasonable to expect a reply from staff to a forum post within 3 hours, let alone to one that is fairly accusatory in tone. Also, many people will downvote and explain their reasoning later when they have more time in the evening.

I will also reiterate my earlier comment that we should establish norms of contacting organisations with such accusations before they are published, unless good reasons exist not do so.

I tried to comment on the page https://ai-risk-discussions.org/perspectives/test-before-deploying, but instead got an error message telling me to use the contact mail.

I really enjoyed this post. Small note: The links in the "If you found this post interesting, you might also like" section are not working.

I intensively skimmed the first suggested article "Technology is not Values Neutral. Ending the reign of nihilistic design", and found the analysis mostly lucid and free of political buzzwords. There's definitely a lot worth engaging with there. Similarly to what you write however, I got a sense of unjustified optimism in the proposed solution, which centers around analyzing second and third order effects of technology during their development. Unfortunately, the article does not appear to acknowledge that predicting such societal effects seems really hard, as evidenced by the observation that people in the past have anecdotically been usually wrong about the sociatal effects of technologies, and that there is no consensus even on the current effects of e.g. social media on society.

To people who found themselves agreeing with this post, I encourage applying it in practice when you actually do encounter first posts (if they obey forum norms). Take this post which was a first post for the author and consider whether people could have been a little kinder with their downvotes. (To be frank, I am a little bitter about my similar experience, so I am a bit sensitive to expressed but not lived norms of welcoming new post writers).

Load more