All of RyanCarey's Comments + Replies

Yes, that's who I meant when I said "those working for the FTX Future Fund"

This is who I thought would be responsible too, along with the CEO of CEA, that they report to, (and those working for the FTX Future Fund, although their conflictedness means they can't give an unbiased evaluation). But since the FTX catastrophe, the community health team has apparently broadened their mandate to include "epistemic health" and "Special Projects", rather than narrowing it to focus just on catastrophic risks to the community, which would seem to make EA less resilient in one regard, than it was before.

Of course I'm not necessarily saying th... (read more)

^In summarising Why They Do It, Will says that usually, that most fraudsters aren't just "bad apples" or doing "cost-benefit analysis" on their risk of being punished. Rather, they fail to "conceptualise what they're doing as fraud". And that may well be true on average, but we know quite a lot about the details of this case, which I believe point us in a different direction.

In this case, the other defendants have said they knew what they're doing was wrong, that they were misappropriating customers' assets, and investing them. That weighs somewhat against... (read more)

There is also the theoretical possibility of disbursing a larger number of $ per hour of staff capacity.

I think you can get closer to dissolving this problem by considering why you're assigning credit. Often, we're assigning some kind of finite financial rewards. 

Imagine that a group of n people have all jointly created $1 of value in the world, and that if any one of them did not participate, there would only be $0 units of value. Clearly, we can't give $1 to all of them, because then we would be paying $n to reward an event that only created $0 of value, which is inefficient. If, however, only the first guy (i=1) is an "agent" that responds to incenti... (read more)

Julia tells me "I would say I listed it as a possible project rather than calling for it exactly."]

It actually was not just neutrally listed as a "possible" project, because it was the fourth bullet point under "Projects and programs we’d like to see" here.

It may not be worth becoming a research lead under many worldviews. 

I'm with you on almost all of your essay, regarding the advantages of a PhD, and the need for more research leads in AIS, but I would raise another kind of issue - there are not very many career options for a research lead in AIS at present. After a PhD, you could pursue:

1. Big RFPs. But most RFPs from large funders have a narrow focus area - currently it tends to be prosaic ML, safety, and mechanistic interpretability. And having to submit to grantmakers' research direction somewhat def

Thanks for engaging with my criticism in a positive way.

Regarding how timely the data ought to be, I don't think live data is necessary at all - it would be sufficient in my view to post updated information every year or two.

I don't think "applied in the last 30 days" is quite the right reference class, however, because by-definition, the averages will ignore all applications that have been waiting for over one month. I think the most useful kind of statistics would:

1. Restrict to applications from n to n+m months ago, where n>=3
2. Make a note of what percent

I had a similar experience with 4 months of wait (uncalibrated grant decision timelines on the website) and unresponsiveness to email with LTFF, and I know a couple of people who had similar problems. I also found it pretty "disrespectful".

Its hard to understand why a) they wouldn't list the empirical grant timelines on their website, and b) why they would have to be so long.

There is an "EA Hotel", which is decently-sized, very intensely EA, and very cheap.

Occasionally it makes sense for people to accept very low cost-of-living situations. But a person's impact is usually a lot higher than their salary. Suppose that a person's salary is x, their impact 10x, and their impact is 1.1 times higher when they live in SF, due to proximity to funders and AI companies. Then you would have to cut costs by 90% to make it worthwhile to live elsewhere. Otherwise, you would essentially be stepping over dollars to pick up dimes.

Of course there are some theoretical reasons for growing fast. But theory only gets you so far, on this issue. Rather, this question depends on whether growing EA is promising currently (I lean against) compared to other projects one could grow. Even if EA looks like the right thing to build, you need to talk to people who have seen EA grow and contract at various rates over the last 15 years, to understand which modes of growth have been healthier, and have contributed to gained capacity, rather than just an increase in raw numbers. In my experience, one ... (read more)

Yes, they were involved in the first, small, iteration of EAG, but their contributions were small compared to the human capital that they consumed. More importantly, they were a high-demand group that caused a lot of people serious psychological damage. For many, it has taken years to recover a sense of normality. They staged a partial takeover of some major EA institutions. They also gaslit the EA community about what they were doing, which confused and distracted decent-sized subsections of the EA communtiy for years.

I watched The Master a couple of mont... (read more)

Interesting point, but why do these people think that climate change is going to cause likely extinction? Again, it's because their thinking is politics-first. Their side of politics is warning of a likely "climate catastrophe", so they have to make that catastrophe as bad as possible - existential.

I think that disagreement about the size of the risks is part of the equation. But it's missing what is, for at least a few of the prominent critics, the main element - people like Timnit, Kate Crawford, Meredith Whittaker are bought in leftie ideologies focuses on things like "bias", "prejudice", and "disproportionate disadvantage". So they see AI as primarily an instrument of oppression. The idea of existential risk cuts against the oppression/justice narrative, in that it could kill everyone equally. So they have to opposite it.

Obviously this is not wha... (read more)

I guess you're right, but even so I'd ask:

• Is it 11 new orgs, or will some of them stick together (perhaps with CEA) when they leave? 
• What about other orgs not on the website, like GovAI and Owain's team? 
• Separately, are any teams going to leave CEA?

Related to (1) is the question: which sponsored projects are definitely being spun out?

Hmm, OK. Back when I met Ilya, about 2018, he was radiating excitement that his next idea would create AGI, and didn't seem sensitive to safety worries. I also thought it was "common knowledge" that his interest in safety increased substantially between 2018-22, and that's why I was unsurprised to see him in charge of superalignment.

Re Elon-Zillis, all I'm saying is that it looked to Sam like the seat would belong to someone loyal to him at the time the seat was created.

You may well be right about D'Angelo and the others.

2. The main thing that I doubt is that Sam knew at the time that he was gifting the board to doomers. Ilya was a loyalist and non-doomer when appointed. Elon was I guess some mix of doomer and loyalist at the start. Given how AIS worries generally increased in SV circles over time, more likely than not some of D'Angelo, Hoffman, and Hurd moved toward the "doomer" pole over time.

Nitpicks:

1. I think Dario and others would've also been involved in setting up the corporate structure
2. Sam never gave the "doomer" faction a near majority. That only happened because 2-3 "non-doomers" left and Ilya flipped.

Causal Foundations is probably 4-8 full-timers, depending on how you count the small-to-medium slices of time from various PhD students. Several of our 2023 outputs seem comparably important to the deception paper: 

• Towards Causal Foundations of Safe AGI, The Alignment Forum - the summary of everything we're doing.
• Characterising Decision Theories with Mechanised Causal Graphs, arXiv - the most formal treatment yet of TDT and UDT, together with CDT and EDT in a shared framework.
• Human Control: Definitions and Algorithms, UAI - a paper arguing that corrig

What if you just pushed it back one month - to late June?

2 - I'm thinking more of the "community of people concerned about AI safety" than EA.

1,3,4- I agree there's uncertainty, disagreement and nuance, but I think if NYT's (summarised) or Nathan's version of events is correct (and they do seem to me to make more sense to me than other existing accounts) then the board look somewhat like "good guys", albeit ones that overplayed their hand, whereas Sam looks somewhat "bad", and I'd bet that over time, more reasonable people will come around to such a view.

It's a disappointing outcome - it currently seems that OpenAI is no more tied to its nonprofit goals than before. A wedge has been driven between the AI safety community and OpenAI staff, and to an extent, Silicon Valley generally.

But in this fiasco, we at least were the good guys! The OpenAI CEO shouldn't control its nonprofit board, or compromise the independence of its members, who were doing broadly the right thing by trying to do research and perform oversight. We have much to learn.

Yeah I think EA just neglects the downside of career whiplash a bit. Another instance is how EA orgs sometimes offer internships where only a tiny fraction of interns will get a job, or hire and then quickly fire staff. In a more ideal world, EA orgs would value rejected & fired applicants much more highly than non-EA orgs, and so low-hit-rate internships, and rapid firing would be much less common in EA than outside.

It looks like, on net, people disagree with my take in the original post. 

I just disagreed with the OP because it's a false dichotomy; we could just agree with the true things that activists believe, and not the false ones, and not go based on vibes. We desire to believe that mech-interp is mere safety-washing iff it is, and so on.

My response to the third link is: https://forum.effectivealtruism.org/posts/i7DWM6zhhPr2ccq35/thoughts-on-ea-post-ftx?commentId=pJMPeYWzpPYNrhbHT 

On the meta-level, anonymously sharing negative psychoanalyses of people you're debating seems like very poor behaviour. 

Now, I'm a huge fan of anonymity. Sometimes, one must criticise some vindictive organisation, or political orthodoxy, and it's needed, to avoid some unjust social consequences.

In other cases, anonymity is inessential. One wants to debate in an aggressive style, while avoiding the just social consequences of doing so. When anonymous users misbehave, we think worse of anonymous users in general. If people always write anonymously, the... (read more)

I'm sorry, but it's not an "overconfident criticism" to accuse FTX of investing stolen money, when this is something that 2-3 of the leaders of FTX have already pled guilty to doing.

This interaction is interesting, but I wasn't aware of it (I've only reread a fraction of Hutch's messages since knowing his identity) so to the extent that your hypothesis involves me having had some psychological reaction to it, it's not credible. 

Moreover, these psychoanalyses don't ring true. I'm in a good headspace, giving FTX hardly any attention. Of course, I am not... (read more)

Creditors are expected by manifold markets to receive only 40c on each dollar that was invested on the platform (I didn't notice this info in the post when I previously viewed it). And, we do know why there is money missing: FTX stole it and invested it in their hedge fund, which gambled away and lost the money.

The annual budgets of Bellingcat and Propublica are in the single-digit millions. (The latter has had negative experiences with EA donations, but is still relevant for sizing up the space.)

It's hard to say, but the International Federation of Journalists has 600k members, so maybe there exists 6M journalists worlwide, of which maybe 10% are investigative journalists (600k IJs). If they are paid like $50k/year, that's $30B used for IJ.

Surely from browsing the internet and newspapers, it's clear than less than 1% (<60k) of journalists are "investigative". And I bet that half of the impact comes from an identifiable 200-2k of them, such as former Pulitzer Prize winners, Propublica, Bellingcat, and a few other venues.

Anthropic is small compared with Google and OpenAI+Microsoft.

I hope this is just cash and not a strategic partnership, because if it is, then it would mean there is now a third major company in the AGI race.

There are also big incentive gradients within longtermism:

• To work on AI experiments rather than AI theory (higher salary, better locations, more job security)
• To work for a grantmaker rather than a grantee (for job security), and
• To work for an AI capabilities company rather than outside (higher salary)

this is something that ends up miles away from 'winding down EA', or EA being 'not a movement'.

To be clear, winding down EA is something I was arguing we shouldn't be doing.

I feel like we're closer to agreement here, but on reflection the details of your plan here don't sum up to 'end EA as a movement' at all.

At a certain point it becomes semantic, but I guess readers can decide, when you put together:

• the changes in sec 11 of the main post
• ideas about splitting into profession-oriented subgroups, and 
• shifting whether we "motivate members re social pres

JWS, do you think EA could work as a professional network of “impact analysts” or “impact engineers” rather than as a “movement”? Ryan, do you have a sense of what that would concretely look like?

Well I'm not sure it makes sense to try to fit all EAs into one professional community that is labelled as such, since we often have quite different jobs and work in quite different fields. My model would be a patchwork of overlapping fields, and a professional network that often extends between them.

It could make sense for there to be a community focused on "effe... (read more)

Roughly yes, with some differences:

• I think the disasters would scale sublinearly
• I'm also worried about Leverage and various other cults and disasters, not just FTX.
• I wouldn't think of the separate communities as "movements" per se. Rather, each cause area would have a professional network of nonprofits and companies.

Basically, why do mid-sized companies usually not spawn cults and socially harm their members like movements like EA and the animal welfare community sometimes do? I think it's because movements by their nature try to motivate members tow... (read more)

Scott's already said what I believe

Yes, I had this exact quote in mind when I said in Sect 5 that "Religions can withstand persecution by totalitarian governments, and some feel just about as strongly about EA."

People would believe them, want to co-ordinate on it. Then they'd want to organise to help make their own ideas more efficient and boom, we're just back to an EA movement all over again.

One of my main theses is supposed to be that people can and should coordinate their activities without acting like a movement.

I still want concern about red

A lot of the comments seem fixated on, and wanting to object to the idea of "reputational collapse" in a way that I find hard to relate to. This wasn't a particularly load-bearing part of my argument, it was only used to argue that the idea that EA is a particularly promising way to get people interested in x-risk has become less plausible. Which was only one of three reasons not to promote EA in order to promote x-risk. Which was only one of many strategic suggestions.

That said, I find it hard not to notice that the reputation of, and enthusiasm for EA ha... (read more)

1 - it's because Sam was publicly claiming that the trading platform and the traders were completely firewalled from one another, and had no special info, as would normally (e.g. in the US) be legally required to make trading fair, but which is impossible if the CEOs are dating

2 - I'm not objecting the spending. It was clear at the time that he was promoting an image of frugality that wasn't accurate. One example here, but there are many more.

3 - A lot of different Alameda people warned some people at the time of the split. For a variety of reasons, I beli... (read more)

See here. Among people who know EA as well as I do, many - perhaps 25% - are about as pessimistic as me, and some of the remainder have conflicts of interest, or have left.

I agree the primary role of EAs here was as victims, and that presumably only a couple of EAs intentionally conspired with Sam. But I wouldn't write it off as just social naivete; I think there was also some negligence in how we boosted him, e.g.:

• Some EAs knew about his relationship with Caroline, which would undermine the public story about FTX<->Alameda relations, but didn't disclose this.
• Some EAs knew that Sam and FTX weren't behaving frugally, which would undermine his public image, but also didn't disclose.
• Despite warnings from early-Alameda peo

There are various reasons to believe that SBF's presence in EA increased the chance that FTX would happen and thrive:

• Only ~10k/10B people are in EA, while they represent ~1/10 of history's worst frauds, giving a risk ratio of about 10^5:1, or 10^7:1, if you focus on an early cohort of EAs. This should give an immediate suspicion that P(FTX thrives | SBF in EA)/P(FTX thrives | SBF not in EA) is very large indeed.
• Sam decided to do ETG due to conversations with EA leaders. 
• EA gave Alameda a large majority of its funding and talent.
• EA gave FTX at least 1-2 of the other leaders of the company.
• ETG was a big part of Sam's public image and source of his reputation.

Ok, it makes sense that a temporary 5x in volume can really mess you up.

I think the core of the issue is that there's unfortunately somewhat of a hierarchy of needs from a grant making org. That you're operating at size, and in diverse areas, with always-open applications, and using part-time staff is impressive, but people will still judge you harshly if you struggling to perform your basic service.

Regarding these basics, we seem to agree that an OpenPhil alternative should accurately represent their evaluation timelines on the website, and should give an updated timeline when the stated grant decision time passes (at least o... (read more)

The level of malfunctioning that is going on here seems severe:

• The two month average presumably includes a lot of easy decisions, not just hard ones.
• The website still says LTFF will respond in 8 weeks (my emphasis)
• The website says they may not respond within an applicant's preferred deadline. But what it should actually say is that LTFF also may not respond within their own self-imposed deadline.
• And then the website should indicate when, statistically, it does actually tend to give a response.
• Moreover, my understanding is that weeks after these self-impose

I agree. Early-career EAs are more likely to need to switch projects, less likely to merit multi-year funding, and have - on average - less need for job-security. Single-year funding seems OK for those cases.

For people and orgs with significant track records, however, it seems hard to justify.

Yes, because when you are at at will employee, the chance that you will still have income in n years tends to be higher than if you had to apply to renew a contract, and you don't need to think about that application. People are typically angry if asked to reapply for their own job, because it implies that their employer might want to terminate them.

I'm focused on how the best altruistic workers should be treated, and if you think that giving them job insecurity would create good incentives, I don't agree. We need the best altruistic workers to be rewarded not just better than the less productive altruists, but also better than those pursuing non-altruistic endeavours. It would be hard to achieve this if they do not have job security.