Will Bradshaw

Karma: 8,359

Will Bradshaw Sep 10, 2023, 4:23 PM
6 points
1 ∶ 0
in reply to: Rockwell’s comment on: An Incomplete List of Things I Think EAs Probably Shouldn’t Do
Thanks for clarifying, I agree category #3 is the most dicey of the three.
How do you see these trials as differing from standard probation? Is it that the chance of a no-hire at the end is higher? Or the length? Or something else?

Will Bradshaw Sep 10, 2023, 2:05 PM
11 points
4 ∶ 0
in reply to: Adrià Garriga Alonso’s comment on: An Incomplete List of Things I Think EAs Probably Shouldn’t Do
In general, I think it is helpful in discussing work trials if people (including the OP) distinguished between three different things that are commonly called work trials:
- Take-home trial tasks / timed online tests, which typically take somewhere in the region of 2-8 hours and are designed to be doable on a weekend or otherwise without work disruption.
- Short (usually 1-3-day) work trials prior to receiving a job offer. This is what I usually think of as being referred to by the term “work trial”. While it’s technically true that these “interruption of regular employment to complete” (quote from OP) I think it’s generally pretty manageable to make this work by taking e.g. 1-2 days of vacation.
- Long (often several months) “work trials” that essentially act as extended paid probation. I’ve seen these at a few EA orgs and sometimes feel iffy about them, but IMO they’re not all that different from standard probation practices in the US. They definitely require “those currently employed full-time [to] leave their existing job”, though.
I’m not sure which of these the OP was including in her claim. Presumably not the first one?

Will Bradshaw Sep 9, 2023, 3:16 PM
6 points
2 ∶ 0
in reply to: Tristan Williams’s comment on: Sharing Information About Nonlinear
One thing I’m confused about: Emerson is the one making threats, so how do I update on the rest of the Nonlinear team?
I was also uncertain about this, but Kat’s comment above seems to indicate (though not outright say) that she supports the threat to sue.

Will Bradshaw Sep 7, 2023, 10:41 PM
17 points
8 ∶ 0
in reply to: Ruby’s comment on: Sharing Information About Nonlinear
Yeah, this seems right.

Will Bradshaw Sep 7, 2023, 10:10 PM
25 points
20 ∶ 3
in reply to: Holly Elmore ⏸️ 🔸’s comment on: Sharing Information About Nonlinear
Most of the allegations in the OP seem comfortably out-of-distribution to me. (Unless the distribution includes FTX & Leverage, but we know how those went.)

Will Bradshaw Sep 7, 2023, 8:38 PM
21 points
7 ∶ 0
in reply to: Ben Pace’s comment on: Sharing Information About Nonlinear
I agree with this, and think it could have been a terrible day for EA if stuff like this surfaced later in a world where Nonlinear had become more influential. But thankfully* we’re not in that world.
(* Thankfully assuming the allegations are broadly true etc etc.)

Will Bradshaw Sep 7, 2023, 6:11 PM
19 points
2 ∶ 0
in reply to: Joel Becker’s comment on: Sharing Information About Nonlinear
Thanks, I think taking the time to make this stronger phrasing publicly is quite valuable (and seems to match what everyone else is saying so far). It’s important that we not engage in guilt-by-association.

Will Bradshaw Sep 7, 2023, 5:43 PM
94 points
51 ∶ 7
in reply to: Geoffrey Miller’s comment on: Sharing Information About Nonlinear
Whatever its legitimate uses, defamation law is also an extremely useful cudgel that bad actors can, and very frequently do, use to protect their reputations from true accusations. The cost in money, time and risk of going through a defamation trial is such that threats of such can very easily intimidate would-be truth-tellers into silence, especially when the people making the threat have a history of retaliation. Making such threats even when the case for defamation seems highly dubious (as here), should shift us toward believing that we are in the defamation-as-unscrupulous-cudgel world, and update our beliefs about Nonlinear accordingly.
Whether or not we should be shocked epistemically that Nonlinear made such threats here, I claim that we should both condemn and punish them for doing so (within the extent of the law), and create a norm that you don’t do that here. I claim this even if Nonlinear’s upcoming rebuttal proves to be very convincing.
I don’t want a community where we need extremely high burdens of proof to publish true bad things about people. That’s bad for everyone (except the offenders), but especially for the vulnerable people who fall prey to the people doing the bad things because they happen not to have access to the relevant rumor mill. It’s also toxic to our overall epistemics as a community, as it predictably and dramatically skews the available evidence we have to form opinions about people.

Will Bradshaw Sep 7, 2023, 4:39 PM
85 points
45 ∶ 3
in reply to: MathiasKB🔸’s comment on: Sharing Information About Nonlinear
I don’t really see the “terrible day for EA” part? Maybe you think Nonlinear is more integral to EA as a whole than I do. To me it seems like an allegation of bad behaviour on the part of a notable but relatively minor actor in the space, that doesn’t seem to particularly reflect a broader pattern.

Will Bradshaw Sep 7, 2023, 1:57 PM
41 points
28 ∶ 5
in reply to: Emerson Spartz’s comment on: Sharing Information About Nonlinear
Could be! I might end up with egg on my face here, in which case I will do my best to admit it. That said, my most important claim is my last: if you wanted me and others to truly withhold judgement, you really shouldn’t have threatened to sue.

Will Bradshaw Sep 7, 2023, 1:27 PM
165 points
67 ∶ 11
on: Sharing Information About Nonlinear
My thoughts, for those who want them:
- I don’t have much sympathy for those demanding a good reason why the post wasn’t delayed. While I’m generally quite pro sharing posts with orgs, I think it’s quite important that this doesn’t give the org the right to delay or prevent the posting. This goes double given the belief of both the author and their witnesses that Nonlinear is not acting in good faith.
- There seem to be enough uncontested/incontestable claims made in this post for me to feel comfortable recommending that junior folks in the community stay away from Nonlinear. These include asking employees to carry out illegal actions they’re not comfortable with, and fairly flagrantly threatening employees with retaliation for saying bad things about them (Kat’s text screenshotted above is pretty blatant here).
- Less confidently, I would be fairly surprised if I come out of the other end of this, having seen Nonlinear’s defence/evidence, and don’t continue to see the expenses-plus-tiny-salary setup as manipulative and unhealthy.
- More confidently than anything on this list, Nonlinear’s threatening to sue Lightcone for Ben’s post is completely unacceptable, decreases my sympathy for them by about 98%, and strongly updates me in the direction that refusing to give in to their requested delay was the right decision. In my view, it is quite a strong update that the negative portrayal of Emerson Spartz in the OP is broadly correct. I don’t think we as a community should tolerate this, and I applaud Lightcone for refusing to give in to such heavy-handed coercion.
What links here?
- Effective Aspersions: How the Nonlinear Investigation Went Wrong by TracingWoodgrains (Dec 19, 2023, 12:00 PM; 350 points)
- Effective Aspersions: How the Nonlinear Investigation Went Wrong by TracingWoodgrains (LessWrong; Dec 19, 2023, 12:00 PM; 188 points)

Will Bradshaw Apr 23, 2023, 8:40 PM
2 points
0 ∶ 0
on: willbradshaw’s Shortform
The world is as it is, now. Tomorrow, it will be a little different. Some of that difference will be because of things you did, or didn’t do.
The differences you make could be inside you, in things you’ve learned or decisions you’ve made. They could be things you’ve written or made, known only to you as yet, but ready to propagate through the world on your command. Or they could already be out there, in the world, jumping from person to person, a chain of effects starting with you but much larger than you, feeding into the great chaotic system that is the world.
What difference that will ultimately make, days or years or centuries from now, is not for you to know. But know that you have the power to make and break worlds. Whatever you do, or don’t do, today, changes the future.

Will Bradshaw Apr 19, 2023, 5:17 PM
13 points
2 ∶ 0
in reply to: Greg_Colbourn ⏸️ ’s comment on: GWWC Reporting Attrition Visualization
Yeah, I also don’t report despite continuing to make my pledge donations. It doesn’t feel like a core part of the pledge to me.

Will Bradshaw Mar 23, 2023, 8:42 PM
1 point
3 ∶ 2
in reply to: Neel Nanda’s comment on: Design changes & the community section (Forum update March 2023)
This comment helped clarify my feelings here. It’s not that the new style is bad, really—it’s unremarkably fine, and after a while I’ll probably stop noticing it. It’s that the old Forum was a really unusually beautiful website, and throwing that away feels quite sad to me.

Will Bradshaw Mar 22, 2023, 4:39 PM
15 points
15 ∶ 1
on: Design changes & the community section (Forum update March 2023)
The functional/structural changes in the redesign seem good. I feel sad about the typographical changes.
The old Forum had a really nice, distinctive “bookish” style that I thought was classy, pleasant to read, and also somehow calming? The new design feels more crowded to me, and also more generic.

Will Bradshaw Mar 3, 2023, 2:58 AM
4 points
2 ∶ 0
in reply to: keller_scholl 🔸’s comment on: keller_scholl’s Shortform
Thanks for bringing this to broader attention. I think I am opposed to this (admitting the journalist), mostly for the reasons you state.

Will Bradshaw Feb 21, 2023, 7:34 PM
222 points
78 ∶ 13
on: A statement and an apology
I feel there’s a bit of a “missing mood” in some of the comments here, so I want to say:
I felt shocked, hurt, and betrayed at reading this. I never expected the Oxford incident to involve someone so central and well-regarded in the community, and certainly not Owen. Other EAs I know who knew Owen and the Oxford scene better are even more deeply hurt and surprised by this. (As other commenters here have already attested, tears have not been uncommon.)
Despite the length and thoughtfulness of the apology, it’s difficult for me to see how someone who was already in a position of power and status in EA—a community many of us see as key to the future of humanity—behaved in a way that seems so inappropriate and destructive. I’m angry not only at the harm that was done to women trying to do good in the world, but also to the health, reputation, and credibility of our community. We deserve better from our leaders.
I really sympathize with all the EAs—especially women—who feel betrayed and undermined by this news. To all of you who’ve had bad experiences like this in EA—I’m really sorry. I hope we can do better. I think we can do better—I think we already have the seeds of something better—but first we need to look hard at what we’re not doing well.
(Also: I think this apology was necessary, and was also unusually well-done, but I disagree with another highly upvoted commenter about it being “healing”. The revelation that Owen Cotton-Barratt did this feels very damaging to my relationship with the UK EA community—which typically does so much better than the Bay Area community on metrics like this! I hope that healing and a degree of rehabilitation can come in time, but I personally feel that talk about healing, redemption, etc, is somewhat premature, and that now is the time to pause for people’s hurt and grief.)

Will Bradshaw Feb 21, 2023, 7:15 PM
25 points
14 ∶ 4
in reply to: Lizka’s comment on: EV UK board statement on Owen’s resignation
It’s now fallen off the front page, at least for me.
Pinning seems pretty important.

Will Bradshaw Feb 21, 2023, 2:24 AM
−5 points
1 ∶ 1
in reply to: Jonas V’s comment on: EV UK board statement on Owen’s resignation
Strongly agree with (both parts of) this.

Will Bradshaw Feb 21, 2023, 2:08 AM
4 points
5 ∶ 1
in reply to: GoodEAGoneBad’s comment on: A statement and an apology
It does seem pretty unfortunate that the last paragraph of the top-level comment was in the same comment as the rest of it.