i don’t want to make claims about the particular case, but im worried if you infer a heuristic and apply it elsewhere it could fail
I think Nonlinear should be assigning enough credence to the possibility that they extremely harmed their employees with a degree of horror and remorse. Not laughter or dismissal.
Sometimes scrupulous/doormat people bend over backwards to twist the facts into some form to make an accuser reasonable or have a point. If you’ve done this, maybe eventually you talked to outside observers or noticed a set of three lies that you had previously been charitable about—and you’re like “hey wait a minute, I totally regret dwelling on feeling bad about this”.
I talked about this with a friend who updated me toward thinking that the scrupulous doormat people are a small enough proportion of the population (or at least certainly they grow out of it) that this is true but a rounding error.
I’ve done this a few times, and I’d like to advocate for people of scrupulous doormat tendencies to have space to not be eternally signaling that they did due diligence before concluding that they dont have deep regrets or behavioral updates to do every time it comes up. People should feel allowed to express themselves aligned with very reasonable needs for catharsis or to reclaim their map of reality. Sometimes this will look like dismissiveness (especially if theyre still bitter about concessions having been extracted).
So if we were applying my point to this case (which, to be clear, we’re not), it would look like “we were horrified and remorseful, went through and did an audit, when we concluded that everything was fine all along we ended up pretty annoyed that we were pressured into doing this unpleasant timesink audit process, and are trying to laugh it off now”—I don’t think it’s necessarily a signal if they don’t prioritize conveying that this process had occurred to you in particular.
(Again re the rounding error thing: very few cases are sufficiently black and white for this comment to apply).
The title isn’t quite right. My current take is that:
There is a ~10k word post sharing some damaging information about someone else.
3-4k words are describing accusations that are disputed.
However, a few have already had substantial evidence provided against them.
Ben doesn’t think the accuracy of those accusations is particularly cruxy for his major takeaways.
Those accusations generally sound pretty bad.
The post doesn’t explain why the (disputed) accusations are salient, if they aren’t cruxy for Ben’s major takeaways.
Ben may not find those accusations cruxy, but I expect many others would find them cruxy for deciding between “there’s some bad stuff in here, but it seems possible that most of the harm was caused by some combination of large cultural differences and poor communication”, and “nope nope nope”.
I think this is even more likely to be true beyond the EA ecosystem. Reputational harm from accusations which later turn out to be false is a real cost. It’s wrong to decide it doesn’t matter because you expect (or want) people in our local circles care about the things you care about, and also be epistemically virtuous in holding uncertainty and rolling back updates when necessary.
The thing I’m noticing is some kind of missing mood. I think that there’s been a failure to inhabit the least convenient possible world, and the general distribution over possible outcomes, and correspondingly attempt to move to the pareto-frontier of outcomes assuming that distribution. This comment in particular seems to be operating in the frame of “You weren’t behaving the way I’d update positively on if the accusations were true”. In the world where the disputed accusations are in fact false, for the specific reasons provided by Nonlinear and others, I think it would be pretty strange to take the course of action that Ben says he would have found sympathetic.
Meta: I’m close with people involved in the Nonlinear accusations. Robert and I were both involved to some degree in the investigation, following along and discussing things with Ben.
To focus on Ben’s comment you called out.
To which you say:
I think Ben might be going a little too far in expecting them to be warning other people of accusations they think are outright false. I think such behavior would be supererogatory and a positive update on Nonlinear, but I think his gist is that they are displaying a severe missing mood here such that is both damning and seems unlikely they’re going to produce anything that changes his conclusions. I agree with that.
I think Nonlinear should be assigning enough credence to the possibility that they extremely harmed their employees with a degree of horror and remorse. Not laughter or dismissal. Unless they are very confident that their employees are lying about both events and the effect on them (I’d be interested to hear a motive), I think a lot of soul searching is in order. That’s what I expect of any compassionate, responsible person to do confronted by claims their employees feel hurt (or the more when your org operation ignores standard safe guareds) . The observed reaction is much more consistent to me with the accused behavior than if I’d they’d said “Oh my god, we are mortified that my employees felt they were so harmed by working for , we’re going to very carefully examine how this came to be and what our part in it was.” That plus cooperation (or cooperative spirit) with a truthseeking here (of which threatening people to end their careers is not compatible, nor suing libel against investgators).
I suppose there’s a thing where I could imagine that Nonlinear team members are pretty “immature” in failing to model an appropriate response to these kind of accusations, and also empathy-wise immature in being more concerned with defending themselves than reflecting on harms they maybe caused, but that same immaturity makes it more likely they were harming their employees without appreciating it.
I am really thoroughly appauled by laughter. My odds ratio is much higher on that in worlds where they caused a lot of harm than in world’s where they’re being falsely accused.
To recap, I think you’re mistaken that’d be pretty strange to do what Ben advised in world’s they’re innocent for reasons they argue. And by extension, I think you’re mistaken on the inference that Ben isn’t considering worlds Nonlinear is false accused and choosing actions that make sense in those worlds too.
I have lots to say on other points, but figured best to tackle one thing a time
I don’t think it would be supererogatory to actively encourage others to speak to someone who you think is either a compulsive liar, or otherwise so detached from reality that their behavior looks like that from your perspective. I think it’d probably be harmful—in that world, what is the good thing that happens as a result of such encouragement?
I’m not interested in going deep on what an ideal response from Nonlinear would have looked like, though I do want to note that this is setting a very high bar which many innocent people would fail to reach.
I don’t think that Ben didn’t do this at all—he obviously did more than nothing when it came to seeking out evidence that might disconfirm some accusations, and hedged some claims. But for the most part, those hedges all struck me as very local to some specific accusations. In another comment, Ben explicitly puts reputational harm to Nonlinear as outside of his circle of concern:
The charitable interpretation of this is that Ben is not concerned about deserved reputational harm. I do not think this interpretation is accurate, based on context (including what this comment is responding to).
I don’t endorse behaving this way when dealing with implacable adversaries, let alone in situations where there’s non-trivial uncertainty. I think that many of the choices[1] about what content to include in the post, and how to contextualize it, were made from this perspective.
Admittedly, I think some of this was exacerbated by a commitment to publish by a certain date.
I find that a strange thing to have said. I wouldn’t have said that.
I think perhaps the actual steelman is that you’ve got two options for how your justice system works:
Option 1 - Private Assessment: Evidence is collected and assessed privately. Public accusations are only made when all reasonable data (e.g. from the accused) has been collected and assessed by high-integrity trustworthy people. The bar for going public is very high. When you do go public, you put anyone accused in a good position to respond (e.g. giving them time to prepare a response.
Option 2 - Communal Assessment: the norm is more free sharing of observations, evidence, both positive and negative. Information flow is encouraged even if it’s critical or damning. You don’t have to have done a deep dive on someone before you’re allowed to say something bad about them that could harm the reputation. Judgments can be formed by a wider range of people earlier, and with more evidence.
I think Ben might be saying believes in Option 2, and that within Option 2, it’s good for him to share information and conclusions he has, because he’s not the final judge. He’s just passing on information to the community (that he’s carefully formed epistemic status around), and the community gets to process that info with other and reach a conclusion.
I do think Ben’s probabilities here play into it. There are costs to putting out false info, but Ben has thought carefully about his confidence and has included that. He reached a high enough confidence in the correctness and importance of it, that he was sure publishing it and starting a public conversation was warranted.
It’s still a bit weird for Ben to say “not my concern”, but I’m actually quite in favor of Option 2 – Communal assessment whereby people more freely spread info, both positive and critical, so we can detect bad behavior and deal with it.
It’s possible to me that 50+% of Nonlinear’s wrongdoing is acting in ways which (if widespread) would make ours a community very vulnerable to harboring abusive behaviors. (Acting in ways such as advocating “we don’t say bad things about you, you don’t say bad things about us”, making statements that juxtapose “don’t say bad things about us” with “we could end your career in a few DM’s”, and the email to us threatening libel suit and damages to the greatest extent possible if we publish.
I do expect a Motte and Bailey where Nonlinear will say “of course you can say negative things, but you should hear both sides out before a public statement and let us prepare a response”, and that’s not crazy, but then the Bailey is a large number of practices and positions that push against people ever saying critical things about you.
And this matters, because if you’ve made it hard for negative information to spread, you make it very hard for anyone to assemble a compelling “justified” that’d meet the high bar for saying anything very critical publicly.
Maybe the ideal is that the employees themselves wrote their own first person accounts, just as evidence, for the public. But I think Nonlinear’s attitudes and behaviors terrified their former employees out of doing that. By luck, I guess, the situation ended up before me/Ben/Lightcone and we took it on, doing a lot of coaxing, confidentiality-promising, etc to get info. But a “pulling teeth” level effort because everyone involved wanted confidentiality.
Getting back to the higher level. While I do think Share Evidence / Communal Assessment will result in some (possible significant) false harms to those accused, I really think it allows for a system that can expose any bad behavior at all without extreme costs. I know you dislike the false positive / false negative framing, but I do think it applies and we are way in the wrong direction of a system that protects abusers from abused. Note that Nonlinear were established, reputable, “powerful” people in EA. Known people. And employees were people in their first EA jobs or so, trying to get into EA. They’re at a significant disadvantage for coming forward (and it took 1.5 years to get this point).
Hmm, I don’t think you need to play that angle that this only bad for young new people vs more established people. Oli has more detailed and robust models here than me, but the charge would be similar issues allowed FTX to do what it did. Norms against being critical/making accusations meant that no one was able to really see what was going on with FTX because even established people in positions of power (e.g. Oli) were stifled by the norms. Though I suppose you could say that FTX, as major funder, had even more power than many established EA figures.
And my claim is if we say that it’s more okay to share critical stuff (both informally/privately) and publicly than people currently treat it, we move into a better community that is not vulnerable to what we’ve seen.
I think one of your points is that it might work if our community has a norm of more freely sharing negative info and then updating in response of other info, etc., but things we discuss (especially online publicly) will have effects on reputation outside of our communities in places where people won’t update on rebuttals, etc.
I think I just want to bite that bullet. I don’t think we can have a functional community if we adopt the level of due diligence, etc. that some people are advocating around this case. I think actually we should make being critical feel substantially lower stakes[1].
Something that might have avoided much of the Nonlinear situation is if the employees had spent more talking with people outside, and gotten encouragement to better advocate for their needs or otherwise leave. If I had a friend who was unhappy with their situation as the employees were, and who told me, I’d have encouraged them. My guess (and this is a guess) is that something about the way Nonlinear operated meant their employees did not feel comfortable complaining to outside people. I could be wrong about this though. But this is part of my case of “no, just encourage people being critical, bad stuff comes from suppression”.
I’m not really thinking in terms of a justice system, though I do think the binary presented is a bit of a false dichotomy—like in everything, there are many axes along which we can make different trade-offs. (And also many axes along which we can make strictly suboptimal decisions.)
I think I basically don’t buy this kind of compartmentalization. Ben does in fact have strong feelings about specific red lines that he feels were crossed (which I largely agree with, conditional on those things having actually happened), and deterring certain kinds of behavior is the main driving force motivating this kind of investigation and publication. If, in the course of his investigation, Ben talked to a random third party who claimed that Nonlinear had conducted a jewelry heist[1], and Nonlinear then provided conclusive evidence that the claim was a total confabulation, would it make sense to include this accusation in the post? I think “obviously no”. There is some judgment being exercised here. There was some such judgment exercised even in the writing of this post! Ben excluded many accusations, both for the sake of brevity, but I think also some that didn’t feel relevant enough to justify their inclusion.
I really do want to stay at the object-level, rather than kicking the conversation up to the meta-level. I think the best way to figure out whether we should be putting in marginally more effort with such efforts, or marginally less, is to discuss specific cases, rather than try to argue about abstract principles and incentive structures. I agree that “we” (as a community) might currently be taking the wrong trade-offs in terms of how much work people put in when deciding to share information like this. That doesn’t speak to whether this case made the appropriate trade-offs rather than overcorrecting, or worse, made unforced errors.
I don’t actually think I’m asking for an increased level of due diligence, strictly speaking. I just can’t figure out what preferences are satisfied by the trade-offs that Ben did make. If he’d skipped soliciting feedback from Nonlinear entirely[2], and written his post with the framing and caveats that would be appropriate for having made that kind of decision, I might have disagreed with either the beliefs that implied those actions, or the values that generated those trade-offs, but I would be less confused. But having decided to run the accusations by Nonlinear (and others speaking in their favor) first, and then receiving moderately surprising evidence in response, it is pretty strange to decide that you’re not going to do much with that evidence. Was there any evidence you could have received, such that you’d substantially change the post after talking to them? If not, why bother[3]?
Selected as an accusation of extremely bad behavior which is largely unrelated to the existing claims.
e.g. because of concerns about potential retaliation, etc.
I model Ben’s response here as being something like “there could’ve been evidence that weighed on issues that were more cruxy to me”, but then this comes back to my disagreement with the truncated circle of concern.
I want to amend/clarify my position that Ben was acting in line my “communal assessment” model of how to approach community justice stuff, and that I think this is better than alternatives.
A core feature of this approach is it being relatively to okay share incomplete incremental information that paint someone in a bad light, even at the risk of missing important context and causing people to update wrongly.
More than just sharing incomplete info, it is appropriate and encouraged to share your epistemic state based on incomplete info. Things are much better if people can more freely share their epistemic states without worrying they haven’t met the bar for doing so (such bars serve to suppress info).
Sharing overall epistemic states and conclusions is valuable because it can aggregate information it isn’t easy or possible to share. E.g. Ben’s overall epistemic state is evidence beyond the contents of the post, because his epistemic state gets to include a lot more information than he was able to impart. Some stories of which I’m aware and are quite damning are not included in the post for some reason, maybe because they were so demeaning to the employees. Obviously the value of hearing Ben’s judgment is tied to what you think of his judgment in general, so some will value it more highly than others.
To make a bolder claim, I think Ben would have been justified in getting to the 75 hour mark, realizing this was not high enough priority to go for another 75 hours, and publish what he’d collected so far (disclosing the process and how he came to believe what he did). I think if I were doing the investigation, I would have published after 20-30 hours with substantially the same accusations and conclusion, but having done less checking of perspective of additional people[1].
I think talking to Nonlinear was supererogatory for sharing, especially in light of suspicions of how they responded. Their actual response (to fully mobilize) I think confirm that giving them more time would have allowed them to make it much harder to get anything out, and so not catering to their requests is even more reasonable than normal.
I can try to do this, though might be hard. Most of my moral reasoning about individual cases routes to thinking about what happens if that behavior was universalized, what incentives it creates, etc. (See moral universalizibilty/ categorical imperative[2]/functional decision theory.
To discuss this specific case though, I think the way it’s gone where the investigation has been immensely costly (dragging out over months, many absolute hours of time from Ben (plus some from me and you, I think it’d be 10+ hours of my time involved with it all)) means that Lightcone will more reluctant to take up future investigation like this.
At the start, when Habryka thought maybe we’d spend 20 hours on it, said he didn’t know it was worth it, because (I summarize possibly unfaithfully) he thinks there are probably ten equally bad situations going on EA and it’s not clear we want to take on that policing role. In a community where you can share harsh accusations without vastly more effort, I think the end result is we don’t get investigations and a lot of bad behavior gets to persist.
(I have questions about CEA related to the Nonlinear accusations, I know they had been aware of them for a long time, and I suspect that the high bar people insist on for sharing anything harsh is part of why we nothing was said very publicly earlier, Nonlinear continued to have a booth at EAG career fair as late as December 2022, etc.
Edit: actually a very good explanation is I think the employees were extremely reluctant to describe their experienced and Ben only got the info he did when the employees had been coaxed/encourage enough to speak up by people close them. That’s a hard barrier for CEA community health to overcome even if they had some awareness of the accusations.
I think the extreme reluctance to speak up is not itself strong evidence. It’s compatible both they have very valid accusations but were intimidated and also compatible with they had weak cases and were reluctant to make them publicly because they expected to get called out/disproved.
)
I disagree that the evidence was moderately surprising in the direction of doubting the accusations. Kat had already stated that she thought Alice was very unreliable, so it wasn’t surprising that was the primary response to the accusations. The texts about the vegan food situation are slightly exonerating but as commenters have pointed out, possibly are damning/confirming of the situation.
But overall I think the evidence Ben got confirmed that he was unlikely to be missing evidence that would substantially change his epistemic state. Three hours is a lot of time to spend with people, and enough to confirm they weren’t likely to produce anything conclusion-flipping.
I mean yeah.
If they’d said “here’s this notebook Alice left behind that describes a conversation between Alice and Chloe about how they’ll gain fame and sympathy within EA by making claims against Nonlinear”, then I’d update.
Or if they said, “we record our hiring interviews and here’s a recording where we very clearly spelled out what the job would be like in detail, setting expectations clear.
Or here’s the detailed accounting showing that room, board, groceries, added up to the salary equivalent of 75k/year
I think you could state that you have that kind of evidence in a three hour call. I think the situation is Ben entered with a lot of confidence in his epistemic state, but thought it was still reasonable to let them raise anything quite surprising. They didn’t (going on the known point of Alice’s accused unreliability), and he made the prediction further time wouldn’t gain much but would cost a bunch, and decided to go ahead.
While Kat has been repeating that ther was no time sensitive, that’s not the case. The investigation has been very taxing on Ben, taxing on the employees, disruptive to Lightcone work (e.g. Ben would often pull other people away from their work to get their opinion), and probably would be even more taxing if Nonlinear was applying all the pressure they could to prevent publication (e.g. threatening law suits, mobilizing their friends to talk to us).
Granted, I had private info I had high confidence in the about the situation before Ben’s investigation started that annoyingly I’m not sure I should even disclose meta details about, but means I don’t think I would have needed as much investigation to reach the same conclusions.
“Act only according to that maxim whereby you can at the same time will that it should become a universal law.”
Is the claim here something like, “you should basically not worry about second-order effects when sharing evidence this way, when it comes to deciding whether or not it’s good to share that evidence (at that moment in time)”? As in, is this a deontological claim, or is there some trade-off and you just want to push much further on the margin than we’re currently at, either because you don’t think the costs are very large (for e.g. missing important context), or because you think the benefits would be significant (in terms of getting a lot more relevant info out there)?
I don’t really disagree with this.
This seems compatible with my previous response:
Re:
I don’t think I buy that it would have made it much harder to get anything out. I agree that it would’ve given them the opportunity to try to impose a bunch of costs on Ben, Lightcone, Alice & Chloe, and maybe others, but not specifically that they would’ve been able to make it substantially more difficult to publish the post. I probably agree with Habryka’s claim that waiting would be inherently costly because e.g. having this in the background would make it very difficult to do productive work on other things.
I think that even if the amount of time spent on the investigation was a mistake, this is compatible with it also having been a mistake to not spend the additional time at the end of the investigation, given the situation at the time.
The surprising part of the evidence was not Kat’s claim that Alice was very unreliable, it was that Kat seemed to actually have corroborating evidence for that claim. Object-level, I think the texts are much more likely in a world where there’s an ongoing two-way failure of communication than in a world where Nonlinear put in basically zero effort into ensuring Alice had food that she could eat.
Maybe, but not cruxy, see previous points re: everybody else in the world.
If memory serves, they claimed to have this, at least for Chloe.
Similarly, I think Drew claimed that he ran the numbers on it, and they came out roughly ok, but both Alice and Chloe were upset by the situation, so he deleted the document. (This is confusing to me.)
Seems like both of these things have in fact already happened, so I think this is basically wrong?
It’s a trade-off and I want to push much further on the margin than we’re currently at. I think the current place we’re at is very costly in the bad behavior it lets accrue and trust it destroys (because people know they don’t have info/stuff is suppressed).
I certainly would like to minimize false accusation that wrongly harm people, but my thought would be in so far as we’re on the false-positve/false-negative tradeoff, I want to push for more false-positives, though I’d like to also make other changes along other dimensions to reduce the cost of those.
(I spoke with Robert offline and he clarified that costs here were things like Nonlinear sending lots of emails and calling a lot, but that you could decide to ignore this and publish anyway.) Some of the response I think they would have done more if it wasn’t published already is to start exert all the social leverage they could (they got one prominent friend to try to persuade Ben against publishing). I think when you’ve got a lot of people already jumping in, wanting you to hear them out, saying they’ll be unhappy/judge you for wanting to publish, it gets very hard to think, you’ll pay a cost for each bid for your attention that gets made, and so on. If not possible to get it out, harder and very stressful.
I’m not surprised at all they’d produce texts that maybe seem to support their account of things.
I don’t know, I think I buy the case that they were willing to put in some effort but they put a pretty steep gradient on getting their help, and it doesn’t reveal failure of communication really. But even if it did, I hold employers (especially in more risky arrangements like living with your employees somewhere far flung[1]) to a high bar for ensuring good communication, and Ben’s post is warranted under “some people worked for Nonlinear and had catastrophic misunderstandings that left them feeling hurt and regretful about having worked there”.
And you think it’s not enough that Nonlinear gets to publish that info themselves in time. Am I reading you correctly in you think that because false accusations are so costly, it’s on a person not just to publish info that’s cruxy to them, but also any information they might collect might be cruxy to others?
I think they can do more of those actions in ways that are costly on the margin. *
The fact they’ve done those already suggests they’d do a lot more.
I’m generally of the opinion that people are allowed to do weird risky things, except when they go wrong, you are especially liable. For example, if an employer normal circumstances (live in Berkeley, have an office) is very unhappy with an employer, I’m willing to be quite charitable with the employer. If the employer isolated the employee, made them financially dependent, lives with them, and the employee then complains of it being the worst experience of their lives, my reaction is “well, you gambled and did this risky thing and you lost, that’s very much a lot on you”.
I agree with holding employers to a higher standard when they engage in non-standard practices, especially so with those that seem risky on the outside-view. I think that in the counterfactual where Ben’s fully-considered beliefs summed up to “some people worked for Nonlinear and had catastrophic misunderstandings that left them feeling hurt and regretful about having worked there”, then I think publishing the post as-is would have been a serious mistake, but I also don’t see how in this counterfactual Ben could’ve had much confidence in that belief without receiving more information.
Not as a general heuristic! My argument is about this specific case, with the combination of factors:
having received fairly strong evidence about at least one accusation was substantially false, this is evidence that other disputed accusations are likely to be false or misleading
having heard from a third-party who seems reasonably trustworthy that the party making many of the disputed accusations seemed to constantly confabulate things
most of disputed accusations not being cruxy for Ben’s bottom line
My current prediction, in the counterfactual world where Ben had decided to wait an additional week, is that Nonlinear would not have invested substantial effort into disrupting the publishing of the post for most of the week, but would have been preoccupied by their own evidence gathering. I think they likely would have attempted costly interventions if, after providing that evidence to Ben, he edited the post and they decided the post still contained (what they considered to be) falsehoods. The reason I believe this is because they did not attempt a large number of costly interventions after Ben’s initial call with them, but only once it became obvious that he was going to publish on Wednesday, whether they’d gotten back to him by then or not.
Flagging some mixture of I need elaboration/confused/disagree here, but let’s focus on below points.
I think this is maybe a crux. I don’t believe he received fairly strong evidence that at least one accusation was substantially false. I think he received evidence that one accusation was more ambiguous/less clear cut and there might be relevant evidence to look at. I think that’s reason to expect that other stories might be less clear cut and the other has a case, but I wouldn’t say it was “fairly strong information that it was substantially false”.
That some of the accusations were stretched/exaggerated, the fact that the many other observations (including from someone that no one had doubted) meant still had a lot of confidence in his models of what happened, to me, mean it was pretty reasonable to stick with that and publish.
I think if it had been, the post likely would have needed to be reexamined to determine the extent of accusations that weren’t well founded (and for which there was evidence). Perhaps this is your epistemic state, in which case how strong the evidence here was is the crux.
I think this is not clear cut.
Is your point, why include them if we have reason to think they’re substantially false and don’t effect the bottom line?
In which case, I think it’s not clear that they’re substantially false (merely disputed and less clear cut, as I believe Ben thinks.
And also as above, I think even if they had been, it’s epistemically very relevant to judging the rest of the accusations if it’s clear at least one person was making verifiably false accusations. Granted, the framing of your post would be different if you’re presenting claims that have convincingly refuted vs not.
Your argument here makes sense and I update on this point.
We saw most of the screenshots responding to the claim that Nonlinear was unwilling to get Alice any vegan food for two days while she was sick with covid. I agree that the screenshots may not be fully exculpatory w.r.t. Nonlinear’s obligations in that situation, but that’s a separate question from whether they’re strong evidence that at least one accusation was substantially false. It doesn’t seem particularly ambiguous to me, but if there’s a joint interpretation of the original claim and the screenshots that seems ambiguous to you, I’m open to updating.
No, I think my point is something closer to what you later say:
That is, I think if I had been in Ben’s position after receiving the evidence we did re: vegan food, I think it would still have been fine to publish the post, even without waiting a week for Nonlinear to provide evidence about the other disputed accusations, if I had contextualized the rest of the disputed accusations with significantly more uncertainy about their accuracy. Ben did do some of this, and I do think it’s asking for quite a lot of work. I also think that putting claims out into the public sphere isn’t a “free” action—people have limited attention, etc. - and so your responsibility increases[1] with the number of claims you advance.
Though probably not linearly.
So I take it what you’re saying is this paragraph doesn’t cut it:
I think that’s a fair thing to say, that he ought to have at least said something more clearly (perhaps at the start saying “Nonlinear disputes many of these claims and says they have evidence they intend to bring. I have seen the evidence on one point and it does make it seem a bit less clear cut. Nonetheless, I am publishing now because....”
If he’d had that, how much better would it have been from your perspective?
I think it’s material that Ben does have a 1366 word section “Conversation with Nonlinear” in which he paraphrases what they said to him, passing along that information.
Moderately? I think one thing that makes it a bit tricky is that many of the accusations were relayed indirectly. If the sections of the post detailing the accusations had mostly been composed of direct quotes from Alice, Chloe, and other third parties, maybe with Ben giving his holistic take/summary at the end of each section, I think they would by default need much less contextualizing, and it would be much easier to examine or contextualize any individual claim.
(Alas, we are out of time to continue this discussion and may not get to wrap it up.)