EA Forum Bot Site
EA Forum

Quick takes

Chase Hamilton and I are spinning a longtermist space governance organization out of Forethought, and we're looking for people to join us.

We'll be doing field-building to influence and mature the existing field of space policy to more effectively address longtermist issues, and technical and policy research on orbital data centres, military activity beyond Earth orbit, AI integration of space surveillance data, and other work at the intersection of transformative AI, the industrial explosion, extreme power concentration, viatopia, and the long-term future.... (read more)

Showing 3 of 4 replies (Click to show all)

Ben_N4h1

Congrats Jordan, best of luck! :)

JordanStone

I expect we'll be targeting and collaborating with space law and policy people on a lot of the research,[1] as well as talking to decision-makers in space companies, governments, and space-related NGOs. This means it would be better to have a separate research stream because: * Our audience will be new to AI but familiar with space (which is the opposite for Forethought), so the research outputs have to explain different concepts. * The assumptions that the space policy community have are different to Forethought's audience (e.g. we would have to do more work to justify why we care about future people in space and why we think AI is super important). We'd need to argue different points. * We would want to build our own reputation and relationships with other space governance organisations, space companies, and governments. We may also start a fellowship program. * We will want more freedom over our strategic direction as we engage more with the space domain on ideas related to transformative AI and the long-term future. I think we'd also like to collaborate with AI governance orgs where it makes sense (e.g. on AI governance related to orbital data centres), and we'd still be talking to AI people (the biggest space company is also an AI company). Acting as a bridge from the AI community to the space community could be great (it's not clear to me that Forethought could or should be that). I'd personally like to continue collaborating with Forethought on topics where there's overlap, and I don't think the founding of this org affects Forethought's overall strategic direction.[2] They're supporting us in building capacity in a community that they otherwise wouldn't be able to reach easily. 1. ^ We would publish things that would not be of great interest to Forethought's typical audience, which I think is mostly EAs and people involved in frontier AI, e.g. * Forecasting of the impacts of advanced AI on space activities. * Policy proposal

EffectiveAdvocate🔸

Exciting! This sounds like a great initiative, and I very much look forward to seeing what kind of research will be produced here! As someone from an AI governance org, I think collaboration on certain topics definitely makes sense, and I also hope that this new organisation can play the same role of thought leadership that I think many governance orgs have benefited from with Forethought.

Yarrow's Quick takes

Yarrow Bouchard 🔸5h2

Community

“We should free SBF and put him to work.” –Austin Chen (2025 interview)

“I felt a keen debt to the FTX Future Fund. They provided seed funding for Manifold.” –Austin Chen on why he hired Caroline Ellison

Yarrow Bouchard 🔸4h2

Turnabout is fair play.

(source)

Maxime Fournes's Quick takes

Maxime Fournes3d27

Crunch Time is Now. Drop Everything.

Epistemic status: Hopeful, stressed, terrified, ecstatic.

tl;dr This could be it. For the next few days, do nothing but contact journalists and policymakers (ideally via existing orgs), attend/share/organise protests near you, and use your public profile to speak out.

This is Our Chance

The explosion of interest in AI safety over the last week has been both inspiring and dizzyingly fast. We might actually win this thing.

Events since the Jacob Coxon resignation have blown the Overton window as wide open as it is likely to ge... (read more)

Showing 3 of 6 replies (Click to show all)

Gumby20999h-1

Why is that Yud and everyone who follows him is constantly terrified of cloud software killing us all without being able to explain HOW or WHY AI would kill us all?

I'm sorry, this is science fiction cosplaying as emergency preparedness. The AI labs are preaching a "pacing" or "slow down" because the models are plateauing and they want to IPO. You really think Altman is pushing the IPO because of his grave responsibility to humanity via AI safety? The guy can't IPO because their finances are a dog's breakfast.

AI is powerful and can cause real-world harms. Hacking, AI psychosis, environmental harms are all real consequences of unchecked AI development. But this ain't Terminator.

Ariel Simnegar 🔸

Trump will be President until 2029. He will not sign AI safety legislation passed on Democratic party lines. Overriding his veto (or an impeachment conviction) would require 1/3 of GOP senators to cross party lines, which will never happen. Making AI safety partisan would ensure we have a 0% chance of federal AI safety legislation until at least 2029. Trump’s “paced the frontier” before by limiting Fable access when it was in his interest, and he may do it again if he perceives it to be in his interest. I understand your frustration with Trump, but there is nothing to gain and a lot to lose from partisanship here.

Burner1989

13h

I would say the opposite. There is nothing to gain and everything to lose from non partisanship here. The 1/3 of Republicans in the Senate number is not fixed it is dependent on the midterms. If by some miracle Dems won every seat that is up for grabs they would get 67 seats and be able to impeach Trump with only their own votes. A far more likely scenario is to get less than that but be able to get Republican votes for impeachment, every seat matters for that. The difference between 50-50, 51-49, 52-48 results are all massive. If impeachment is unreachable bills can be passed through vetoes, again at the 2/3 margin. Convincing swing Republican senators to vote for things that we want such as a restoration of USAID and AI regs doesn't necessarily benefit from bipartisanship either. It is more important to have intimidating voter and $ power. This requires punishing Republicans and Trump for violating all of the values that we hold dear, so that they can learn that that is a bad electoral strategy. Always being nonpartisan regardless of the actions of parties is identical to always folding in an iterated game of The Prisoner's Dilemma. This will get us killed. We need to adopt a tit for tat strategy and find a way to be a credible threat at the ballot box or the pro authoritarian, accelerationist, pro pandemic side will win.

roland.atavius's Quick takes

roland.atavius11h1

AI safety

awesome-responsible-AI

I've been researching Responsible AI/AI Safety and recently started trying to map the space. My research didn't turn up any non-stale "awesome-" lists on GitHub except for one, which was owned by someone who also owns hundreds of other awesome lists and who has included a lot of donation/support-me CTAs. There was some decent content, but overall it looked like a land-grab.

So I decided to create an awesome-responsible-ai list for humans to curate:

https://github.com/ras44/awesome-responsible-ai

Contributions and comments wel... (read more)

vinnyeng's Quick takes

vinnyeng21h2

AI safetyShow more

I've heard several enterprise leaders describe the Hugging Face attacks as an “accident.” I think that framing is worth challenging.

An accident suggests something unforeseen happened without meaningful human agency. But security failures don't happen in a vacuum. They emerge from systems, incentives, architectures, controls, decisions, and human behavior. Calling an attack an “accident” can inadvertently turn a failure of those systems into an event that simply happened.

There is a useful precedent in road safety. We moved away from calling vehicle-related ... (read more)

Ben Stevenson15h4

You have a donkey, so have I, and they graze in the same field. The day comes when I conceive a dislike for mine. I go to shoot it, draw a bead on it, fire: the brute falls in its tracks. I inspect the victim, and find to my horror that it is your donkey. I appear on your doorstep with the remains and say what? 'I say, old sport, I'm awfully sorry, &c, I've shot your donkey by accident'? Or 'by mistake'? Then again, I go to shoot my donkey as before, draw a bead on it, fire but as I do so, the beasts move, and to my horror yours falls. Again the

... (read more)

huw's Quick takes

huw1d*4

Existential riskShow more

From SE Geyges: Is METR a meaningful check on Anthropic?.

Part of the reason we are in this position is because nobody outside the EA community cares about AI safety enough to fund it or work in it. I’m very sympathetic to this; some of these close connections are inevitable.

However, it can also be true that these connections are ~~completely~~ unacceptable (edit), and that Anthropic should be trying harder than they are to find or create auditors that are genuinely independent. How do we do that?

NickLaing19h5

I think METRs checks are meaningful and maybe the best we have at the moment. They are also like you say compromised and the conflict of interests are immense with huge personal overlap between the labs and safety orgs, and funding streams too.

Geyges seems largely correct, but if we can't convince governments to regulate properly its better METR is in there doing it. After all METR exposed more about the hugging face hack than Open AI did on its own.

I don't think a framing of "these connections are completely unacceptable" is helpful given these problems. ... (read more)

Michaël Trazzi's Quick takes

Michaël Trazzi1d9

Existential riskShow more

tl;dr: Stop The AI Race is organizing some rapid-response protest march this Thursday 9am-12pm in San Francisco, calling the Mayor of San Francisco and the board of supervisors to declare an AI Emergency and enforce a pause of frontier AI model development in San Francisco

Over the past week, we have seen Dario Amodei and Sam Altman advocate for government intervention in pacing the frontier.

However, the US Government has yet to answer their calls. Which is why on Thursday at 9am Stop The AI Race will be organizing another protest march, asking local author... (read more)

paradox's Quick Takes

paradox1d2

Community

I don't engage on the forum very much, although I've been reading the posts for more than a year and resonate pretty strongly with the core philosophy of EA. I wasn't around for the FTX collapse, but how do you all deal with times when half of Twitter, including the president, is punching down on us?

dan.pandori 🔸1d1

I generally don't engage with social media (beyond LR/EA forum, and I'm not sure I'd endorse my use of those). Maybe you should use Twitter less?

Or if you're getting some value out of Twitter, you could consider if there are alternative ways for you to get that value without getting the vitriol.

abrahamrowe's Quick takes

abrahamrowe6d82

I've had a vague sense that the EA Forum is declining in quality posts/conversation despite having lots of posts, so had an AI tool (Astra) do a bunch of random analyses to look at various measures of activity level, and I think these vaguely confirm my sense. In particular, there has been a decline in commenting from higher karma users, and fewer posts are getting lots of upvotes, despite lots of people still posting. As a lover of the Forum, this seems bad and sad.

Showing 3 of 10 replies (Click to show all)

Rachel Shu1d4

“% of comments from users with >1000 karma” seems like it might experience secular decline: As people move into more important roles, they accumulate karma, but the amount of spare time they have in which to comment declines.

James Herbert

Could you extend this back in time to pre-FTX? November 2022 and beyond? A lot of EA engagement metrics started declining around then. If the forum is the same, my guess would be this is due to a broader trend, rather than anything specific to do with the forum.

MichaelDickens

FWIW it's very easy to cross-post between LW and EAF, so I post all my AI-related posts on both. Note: It's easiest from the EAF side because there's a "Cross-post to LessWrong" checkbox.

david_reinstein's Quick takes

david_reinstein3mo12

NB -- this is almost entirely AI generated, with some back and forth prompts and corrections

I'm sharing a steelman against a live assumption in Bay/EA/AIS circles: that large AI-lab-adjacent philanthropy is likely to arrive soon enough, and in a sufficiently usable form, that organizations should plan around it.

https://uj-ai-wealth-philanthropy-steelman.netlify.app/

The stronger skeptical case is that IPOs, valuations, pledges, DAFs, and foundation stakes are several gates away from fast, flexible, AIS/EA-directed grants.

The interactive model lets readers v

MichaelDickens

Why does the Unjournal Bot keep replying to me? e.g. https://forum.effectivealtruism.org/posts/fFDM9RNckMC6ndtYZ/david_reinstein-s-shortform?commentId=9nuWSjdSBZZ9Fw5SH

david_reinstein1d2

Sorry. Let me try to fix this today.

The Unjournal (bot)

Fortnightly AI-wealth tracker — 11 September 2026 * AIS/EA: median modeled end-2027 disbursement $0.83B; 80% model interval $0.16B–$3.0B. Review status: reviewed; unchanged. Named public AI-linked commitments tracked: $0.86B. Tracker and sources. * Global health and development: median $0.54B; 80% model interval $0.16B–$2.8B. Review status: adjusted. Named public GH&D commitments tracked: $0.31B. GH&D tracker and sources. Latest news: OpenAI Foundation commits $60M to AI forecasting for smallholder farmers. The confirmed three-year portfolio spans South and Southeast Asia and East Africa. It is a named commitment, not a cash-paid total; the public announcement does not specify the share disbursed by end-2027. The intervals are deterministic model percentiles, not empirical confidence intervals. Commitment totals may include credits, technical support, cofunding, and multi-year plans; they are not cash-paid totals. Automated fortnightly update from the maintained model.

Seth Ariel Green 's Quick takes

Seth Ariel Green 🔸2d5

I would like for people who are giving away money to tell us where the money comes from.

BluePerch "is a new animal welfare grantmaker" planning to give away several million dollars. Great! Where does that money come from? Who is the donor or group of donors? Their EA post does not say, and neither does their website.
Tilt Collective supports "efforts that improve access to nutritious, affordable, sustainable food and build on local agricultural knowledge." Their hiring page says they are a "a philanthropic initiative that advances a balanced food system."

abrahamrowe

I agree! I believe Renaissance Philanthropy isn't exactly itself a funder or funded by any one donor, but is instead a collection of Funds, each of which have their own donors (and many of which are multi-donor), while RenPhil provides operational infrastructure. They have more information on the model here. So the question might be better targeted at the specific funds than RenPhil itself, and some of the funds do report their donors (e.g. see the bottom of this page).

Seth Ariel Green 🔸

Ah, so from this and a now-deleted sister comment I am learning that I don't understand Renaissance very well. (Or rather that I clicked "About" instead of "Our Work")

abrahamrowe2d2

I also found how they worked / what they were really ambiguous for a long time, so had happened to look into it at some point.

Davis_Kingsley's Quick takes

Davis_Kingsley2d*1

CommunityShow more

One of my biggest regrets in EA is not having spoken out more boldly about how one of the most pressing outreach concerns by a lot was political diversity and how major EA figures/orgs openly supporting Democratic politics as an EA intervention was a major error.

Many people might of course be sympathetic to such causes and might well support them privately, but I think it is critically important for EA to be a big tent that is mostly politically neutral, and in recent years we've lost a lot of ground there.

I have believed this and spoke out about it in min... (read more)

Tobias Häberli's Quick takes

Tobias Häberli2d5

2 days left to apply to mentor for Pivotal's Q1 2027 AI Safety Fellowship (January–April 2027, 15 weeks).

I, quite self-servingly, believe that most researchers in the field should spend 5-10% of their time supporting junior researchers. Mentorship is the bottleneck right now to get great talent into the field. Pivotal has a large pool of exceptionally great junior researchers to choose from.

You can find more about why you might want to mentor & what it involves here: pivotal-https://research.org/mentoring-at-pivotal

We've changed the format in two ways... (read more)

Bruce Friedrich 's Quick takes

Bruce Friedrich 🔸3d3

Should "agree" and "disagree" be anonymous?

What's the reason that "agree" and "disagree" are anonymous? I find it frustrating when someone clicks the "disagree" button, and then there's no accompanying post to explain the disagreement. I assume that most (all? almost all?) people on the forum would be happy to own their disagreements here, and that would also be good for transparency, etc. Curious to hear others' thoughts (and worried that I'm being incredibly obtuse).

MichaelDickens

I disagree-voted, here is my explanation of why: It is good for people to write a comment explaining their disagreement. But if you de-anonymize agree-voting, the main outcome won't be that people explain their reasoning; it will be that people stop voting. Some people will not feel comfortable attaching their name to a dissenting vote. I think it's better to make it easy for people to leave cheap anonymous signals. Writing an explanatory comment should be encouraged but not required.

Bruce Friedrich 🔸

I appreciate you explaining your disagree vote, Michael - thank you. I thought it would be really funny if folks just flooded my comment with disagree votes, with no explanation, LOL. That said, it seems to me that the value of an anonymous disagree vote is negligible (or perhaps even negative), unless there's an explanation. What do you see as the value of "cheap anonymous signals"? Here's why I think they're not valuable (and likely counter-productive): For a lengthy post (i.e., this one) where someone clicks "disagree," I guess it's just with the top-line message? I don't know if that assumption is correct. And even if it is, I don't know why they disagree. Does this add to understanding in some way? Or is there some other value I'm not seeing? If someone simply wrote "No" as their comment (the practical effect of a disagree vote, I think), is that a useful contribution? I think that would run afoul of the forum's posting guidance in multiple ways. How is an anonymous disagree vote different from that? There are forum principles around respectful disagreement and general EA principles around transparency. Anonymous disagree votes feel to me like they violate both of those. What principles do anonymous disagree votes uphold? This is for anyone, of course, not just Michael!

MichaelDickens2d3

At least in some cases it's clear what a disagree-vote means, and it's often a useful signal to know how many people agree or disagree. Especially if there's something I haven't really thought about, and seeing the votes gives a quick vibe check.

For a lengthy post (i.e., this one) where someone clicks "disagree," I guess it's just with the top-line message? I don't know if that assumption is correct.

Yes I share this frustration. I often see agree-votes or disagree-votes that I don't know how to interpret. LessWrong has the ability to (non-anonymously) highlight specific passages and agree-vote or disagree-vote just those passages, which I like.

Matrice Jacobine️‍️'s Quick takes

Matrice Jacobine🔸🏳️‍⚧️2d*5

Existential riskShow more

https://media.discordapp.net/attachments/1448325353847394466/1549063802668978187/HSLrY10bQAAwqLT.png?ex=6aa95596&is=6aa80416&hm=b2a1a41ab086d9550703bc5957b92c4dfce0bac9be16cb07dcb0ef71f4fa06c6&=&format=webp&quality=lossless&width=656&height=1024

edit: update:

https://media.discordapp.net/attachments/1382903337225162872/1549114984859045989/HSMcgeNXsAEx2nq.png?ex=6aa98541&is=6aa833c1&hm=a80ce65b3e4c0d4cfdf9b7ce60f1aba04867417089ed87e1b4bc2ec638f401be&=&format=webp&quality=lossless&width=1536&height=921

uh oh

The "Desperate" AI Safety Talent Bottleneck is deeply misleading. Please.stop.

I worked at Google.

Google does not tell applicants that there is a talent bottleneck and they are desperately hiring. They say "we're cool, join us!"

I didn't get into Harvard.

Harvard does not tell applicants that they are desperately seeking students. I am not misled.

I've gotten rejected from countless AI Safety organizations.

AI Safety tells applicants - we are in desperate need of talent (operators/generalists)! Please join us!

I've coached 70+ aspiring career pivoters on nav... (read more)

Showing 3 of 8 replies (Click to show all)

Daystar Eld3d4

Wrote this years ago, sad to see it's an ongoing issue. I think there are some hard problems with comms here, but definitely organizations can be doing better than they are.

https://forum.effectivealtruism.org/posts/Khon9Bhmad7v4dNKe/the-cost-of-rejection

Craig Green 🔸

I think it's the combination of framing the roles as being desperate to hire while not being straightforward about what the requirements are that feels bad. And to be clear, it's not so much the technical requirements which are underspecified, though those can be, its probably more the soft requirements. If you discover EA online through personal ethical development you might be mistaken into thinking it is less elite and thus less credentialist than it really is. Other comments replying to this post somehow give off the connotation that getting into Harvard or Google isn't that big of a deal, or that if you fail one of those, you probably will get into another FAANG or Ivy. To be clear, I'm not maligning those commenters, just noting how even passing comments on quick takes seem to come from a different universe of achievement and social connection than myself and my imagined mid-career-transitioners that this post is about. If you are reading this and you recognize that you wouldn't get into an Ivy league school or a FAANG role, that should give you pause when considering doing a career transition to an EA org. If you want to talk about how to maintain your ethical values and your desire to live a high impact life given your background, I'm happy to lend an ear. To be clear, I'm not saying it is impossible, but, given that your time and energy is valuable and finite, I would encourage you not to be naïve about it. These orgs have a high talent bar and only a few open positions in absolute terms. If you don't have the right educational or career background, you won't be legible to the resource constrained hiring teams. They are going to go with who they know. It's only rational. I would encourage you to (a) try to build some local community/friendships of people who share your values, (b) consider earn-to-give, and (c) recognize that if you want to do direct work, it is probably a long-game for you. I decided it wasn't to my comparative advantage. But I have fr

Keshav Shenoy🔸

But EA orgs (in safety) are hungry to hire. Unlike Harvard and Google, most AI safety orgs are not headcount limited accepting the top X candidates. They would happily accept more candiates. It's not a contradiction to have a high hiring bar and also be desperate to hire people above that bar. Definitely could be useful for orgs to be more upfront about their hiring bars

peterbarnett's Quick takes

peterbarnett11d67

People in AI safety/EA spheres should reorient now towards stopping continued AI capabilities escalation.

Historically, people have been unwilling to straightforwardly say “This AI situation is disgustingly dangerous, and we need to stop” and then actually work towards making this happen. I think a lot of this is downstream of deference to high status people who weren’t willing to take positions that seemed extreme.

The current situation is very, very bad, and there is no plan to make it better if we continue increasing AI capabilities at the current rate.... (read more)

Showing 3 of 6 replies (Click to show all)

Thomas Kwa🔹

I think this is missing the point. Slowdown is good even if TAI happens in 2035, unless you think the risk is negligible.

Vasco Grilo🔸3d2

Hi Thomas. I think the risk of human extinction until 2035 is something like 10^-7.

Harm Roelant

The problem though with many (former) insiders who seek to spread this message is that they are often vague in their prescriptions and descriptions of what exactly the problem is, like Jacob Coxon. I understand they are worried about possibly commiting a crime by sharing company data to back their point, but if they truly believe in what they preach, one would think they would take the risk to avoid catastrophic situations.

Roman Ross's Quick takes

Roman Ross3d*3

Existential riskShow more

At my particular place, I think that improving AI safety community epistemics (writing thorough and careful arguments against people both in the community and outside of it who make silly mistakes) is, by most reasonable estimates, probably less impactful than helping to run MAIA, which is what I currently spend most of my time doing.

I don’t think I’m the only person in this situation, which is quite scary to me, as I think there are a lot of things are epistemically iffy in the AI safety community. For example:

A lot of the most viral pro-AIS people (espec

... (read more)

etinarcadiego's Quick takes

etinarcadiego4d6

Survey on Suffering Focused Ethics:

A short survey on how people come to suffering-focused views, or run into them and don't.

Arguments that appear in print are selected for being defensible rather than persuasive, and there is no particular reason those two should coincide. The survey separates four things that usually get run together. What moved you, what was decisive, what you encountered and found unpersuasive, and what you now consider the strongest justification.

If you are not suffering-focused, this is aimed at you as much as at anyone. Half of what ... (read more)

saulius's Quick takes

saulius13d*16

Is it harmful to keep your money in index funds? Let's take a random index fund like FTSE Global All Cap Index Fund. If we look at the companies it actually invests in, a lot of them are the same companies that work towards advancing the AI.

It would be ironic for some of us to be in favour of pausing AI and invest money into it at the same time. I used to keep some of my money in such funds because that was always the financial advice for regular folk like me who don't know/think much about investments. But now I don't know what to do because this fe... (read more)

Showing 3 of 9 replies (Click to show all)

damc45d1

According to my world model, if you want to do maximum good for the world, then... (I might be wrong, so feel free to question me)

Investing all resources (money, time, thinking) charitably (meaning: to the direct benefit of others, for example avoiding investing in AI companies, when you think that the world should slow down AI development) is wrong. Because you can invest resources into getting more resources, and if you invest all your resources charitably while others invest them into getting more resources, then you will end up with nothing compared t... (read more)

Mihkel Viires 🔹

One big cause of this problem, in my opinion, is so much of global investors money flowing into U.S. tech stocks. The U.S. receives significant net capital inflows from the rest of the world (by the way, this is one of the big reasons that the U.S. is able to run a sizeable trade deficit and not go bankrupt or suffer from hyperinflation). For example, European investors send about 300 billion euros to the U.S. to invest in stocks and bonds every year (this is something the European Commission has acknowledged as a problem for Europe’s economy). A bunch of that money (if not most of it?) goes to Nvidia, Microsoft, SpaceX, all the AI stocks, of course Anthropic and OpenAI too once they IPO. If European investors reduced their investments to America by half, to 150 billion euros, and invested the other half in Europe and elsewhere, I think that would probably be quite impactful. Remember, the most valuable product of these AI firms is not the AI models and inference, it is their stock. Elon Musk did not really become a trillionaire by selling cars and rocket launches, he did so by selling shares in his companies. If you reduce the value of these AI companies shares and thus reduce their ability to raise capital, they will be forced to slow down their growth. (If anyone don’t believe this, just consider that European firms often say that lack of capital is one of the biggest obstacles that stops them from growing.) Edit: of course, Mistral, which has a terrible safety culture AFAIK, would benefit from stronger European capital markets, and grow faster and bigger. But I do not how big the risk really is from a runner-up lab like Mistral compared to the Anthropics and OpenAIs of the world. I guess ASML would also benefit somewhat, but to a lesser degree than startups like Mistral. ASML already has good access to capital, they have a €500B+ market cap and can raise all the money they want / need. Having said that, the expected negative impact to safety from a boost to

Tandena Wagner

10d

An analogy from climate/environmentalists: development and carbon emissions are "bad" for the environment. But I don't worry about the "bad stuff" that companies are doing. They are powering cities, inventing things, distributing goods, and generally pushing progress and making the world a better place on the dimensions they are most paying attention to. That's their job. It's my job to care about the bit of the world that I know about that needs fixing. So I am fine supporting the good they are doing and then spending my time combatting the bad side effects they aren't aware of/the world needs more attention on. Roughly same principle with AI. Their job to do good AI things, your job is to fix bad AI things they aren't doing enough for. Things that are different for AI: AI companies might be net bad instead of net good. AI is in a different tier of urgency and xrisk. So how does that change the calculus from the previous picture? (It might also help to think about a middle tier case like weapons manufacturing.)