We're entering the realm of pure speculation here, and I think we're starting with wildly different priors and have insufficient evidence atm to update them. I feel your suspicions are overtly cynical and are inconsistent with what I feel is the best model of Caroline's personality from what can be reconstructed from all the info in the public domain. Then again, you could retort that she has been involved in a billion-dollar fraud and that it's prudential to be suspicious. So we're stuck in a rut, for now.

Just anecdotally, I'd like to mention some words of Judge Kaplan about her during the trial and sentencing: “I've seen a lot of cooperators in 30 years here. I've never seen one quite like Ms. Ellison”. He was particularly impressed by the accuracy and consistency of her testimony and by her exemplary collaboration, and truthfulness, even against her own interests: “I don't remember a single time when she was caught in the slightest error of fact, the slightest inconsistency”. He definitely considered her remorse genuine, and I can easily imagine that a judge of his experience is a better judge of character than me.

Not in the public record, and I have no privileged information, so the "after prison" is indeed unattested. I would say, though, purely as a personal belief, that it feels very likely that she would have continued to do so afterwards and before she started with Manifund.

The main source is Caroline Ellison's September 2024 sentencing memorandum and the supporting letters submitted to the federal court by her lawyers. There's also contemporary reporting by The New York Times (which touches on her rejection from a tax-preparation position), The Wall Street Journal and other newspapers that likely draw on those documents and other sources .

Good. I'm trying/will keep trying to do my part with this. I think a lot of EAs, though (perhaps some of the leadership?) actually consider the current turn of events as positive, from a model of the type: this gives us visibility and positions us as anti-Trump/Maga at a time when Republicans will likely lose midterms and possibly the next election cycle; plus, this makes us more visible and attractive to a prospective audience of young, left-inclined people who are a better target audience than people on the right.

Fair I guess. Still, I feel you are undervaluing how difficult it likely is for Caroline to find, right now, any job that matches her skills and talents and/or could be impactful and aligned with her values.

I think I agree with the assessment, but what proposals have you thought about to stem this tide?

I think you are making two assumptions here. Please correct me if I'm wrong or misinterpreting you.

1) That Caroline's role at Manifund involves making decisions about how significant philanthropic funds are allocated.

I agree that the job description quoted from Caroline's text ("continue improving our technical infrastructure, working on operations and customer support, and posting updates on our newsletter") may not cover all her responsibilities at Manifund. As you point out, Austin starts his post by saying that Caroline has accepted a role at Manifund "to develop our funding platform and research how to effectively direct philanthropic dollars".

I don't dispute that this is a potentially important role, particularly given that she's apparently Manifund's second full-time employee. And I agree that her research and recommendations could influence funding decisions, perhaps substantially. But I think you're going beyond what the announcement actually establishes when you describe her job as one involving "making impactful decisions about how not-insignificant quantities of money are directed". Researching how to direct philanthropic funds is not the same as exercising decision-making authority over their allocation.

Do we have any evidence that Caroline will actually be making those decisions, approving grants, or exercising control over donor funds? Has anyone asked Austin about this, if it is their main bone of contention? I think not. I also think those are important distinctions, particularly when discussing the responsibilities it would be appropriate to entrust to someone with Caroline's history.

2) That there are other paid roles in the EA ecosystem that Caroline could reasonably and effectively fill.

I agree that there are probably many roles for which Caroline has the necessary skills and qualifications, vut how many EA organizations would actually be willing to offer her one? Do you know of any organization in the EA ecosystem that would be, or would have been, willing to give Caroline a chance, besides Manifund? Myself, I think the number is likely to be very small, and I would be happy to stand corrected.

The distinction matters because there is a considerable difference between saying that someone is qualified for a position and saying that the position is actually available to them. If we're going to argue that Caroline should be rebuilding her professional life elsewhere, I think it's reasonable to ask whether those alternative opportunities genuinely exist.

Finally, I think I agree that it would have been preferable, all else being equal, if the organization she had joined had no unresolved legal or financial issues connected with FTX. However, I don't think the mere existence of litigation or negotiations with the FTX estate establishes wrongdoing or bad faith on Manifund's part. 

Caroline hasn't done the work to earn that trust back. Moreover, her willingness to take on a role she should have known would cause harm to others makes me concerned that she is not on track to earn that trust back anytime soon. And Manifund hasn't earned the trust that it will be a responsible steward of the community's trust on her behalf. This was a unilateral call by Austin and Caroline, and I think it traded down our collective trust.

All my sympathy for the FTX harm you had to live and suffer up close. I don't doubt how harrowing, stressful and painful it must have been for all who suffered it, for all the careers that were affected and for the overall damage that was done to the EA brand and its capacity to 'do good better'.  All my sympathy as well for the challenges of avoiding bad publicity being piled upon EA, all the more so at a time when the US administration seems rather keen on maligning the movement. I would say, though, that you do seem to be presenting a bit of a catch-22 scenario here. How is Caroline supposed to 'earn that trust back'? As far as I've known (which is to say, what is accessible through the public record) she has been engaging since the time of the trial and before and after prison in many iniciatives and work of the kind that is restorative and good for self and others, even if not very impactful: volunteering, soup kitchens, teaching adult-literacy classes, volunteering as a tax preparer for low-income people, helping to send books to prisoners... and she's had great difficulties in finding ordinary paid work (even been asked to leave in some of the work just mentioned when her employers discovered who she was). It's been four years since the FTX debacle. What metric for 'earning back trust' would you consider reasonable? A decade or more doing work like this, out of the public eye, and completely dissociated from EA activities? Would it be okay if she were working in a private charity that is disconnected from EA? Is she supposed not to work in anything related to her skills and talents, even when it doesn't involve investing of safeguarding investments?. As Lorenzo Buonnano says, "improving technical infrastructure, operations and customer support, and posting newsletter updates in a low-paid role at a small charity" feels like a decent fit for a person with her skills and talents, and a desire to do good and show her worth.

Perhaps I am missing some important context too. Some is likely pretty easy to spot: I didn't get interested in EA until after FTX, so I didn't experience either personally or indirectly most of the fallback and the negative consequences for people in the space. I am also the type of person who is very indifferent to PR considerations, and I lean much closer to Rationalist-flavored EA than the other flavor. And I am very positively biased towards Caroline, given the mental map I've made of her from reading about her, her blog and from all the minutiae of the FTX trial. Be that as it may, I find all the arguments you make here pretty rationally convincing (but then I would, wouldn't I?).

Perhaps my wording was too broad. What I meant by “strict deontological behaviours and never be willing to engage or condone even very mildly greyish-white stuff” was specifically in reference to the failure modes that brought her to her current situation. I would expect her now to be exceptionally reluctant to lie, conceal material facts, cross clear legal or fiduciary boundaries, let consequentialist reasoning talk her into something she independently thinks is wrong, or defer to a charismatic authority figure against her own judgement.

I can understand why many EAs would be extremely wary of the reputational consequences of having her work in anything associated with EA. But from her point of view, assuming she is a well-intentioned person who wants both to do as much good as she can and, perhaps, compensate in some measure for the enormous harm she helped cause, I don't think it would make much sense to adopt a rule of avoiding any high-impact work for which she is unusually well suited merely because her involvement will itself be controversial. And, as this thread illustrates, her options for that kind of work are presumably much narrower than they otherwise would be.

I also don't think the Manifest example is very probative here. EA isn't ideologically homogeneous. There is, at least in my mental model, a substantial divide between a more left-oriented tendency that places greater weight on movement growth, reputational considerations, socially acceptable causes and “Classic EA™”, and a more Rationalist-influenced tendency that places greater weight on radical truth-seeking, willingness to explore weird or unpopular ideas and interventions, quantitative arguments even when their conclusions are socially uncomfortable, and comparatively more attention to AI and longtermism.

You can certainly disagree with the latter tendency, and particular decisions made under that banner can of course be bad. But merely inviting heterodox or objectionable people to a conference is not, to my mind, evidence of the kind of ethical laxity that would make me worry about Caroline repeating her Alameda failure mode. “Works for a company that makes controversial judgement calls about whom to platform” and “is willing to lie, commit fraud, or override serious moral reservations because a charismatic superior has produced a consequentialist justification” are very different things!