Toby Tremlett🔹

An­nounc­ing: Clue­less­ness Cri­tiques Competition

• Toby Tremlett🔹 19 Jun 2026 8:22 UTC

  2 points

  0 ∶ 0

  on: Pre­dictable Up­dat­ing About Fund­ing In EA

  I’m curating this because I think it’s a valuable and sober reflection on the (possible, future) funding situation. It’s (always) hard to reason about a big event that might not happen, or might not happen the way you think, and I appreciate Sam’s more bayesian approach here.

  Some other great posts on the (possible, future) funding situation, if you’d like to read more:

  • Reflections on Anthropic and EA

  • Incoming money, integrity, and collective action problems

• Toby Tremlett🔹 17 Jun 2026 11:28 UTC

  3 points

  0 ∶ 0

  in reply to: Elias Schmied’s comment on: A sim­ple ar­gu­ment for try­ing less hard

  “I’m not sure why we should expect the group to be well-calibrated when no individual is?”—Something something marketplace of ideas? An analogy is a court, where the prosecution and the defence both have their conclusions assigned to them beforehand. They are both epistemically vicious, in opposite ways. Then the idea is that the best arguments win on their merits. I’m not sure quite how this analogy would fit though (I’ve had false starts writing a blog post on this a few times).
  
  “I certainly don’t endorse doing nothing” Yep I get that this argument is only one consideration, but my point extends to “trying less hard” as well I think.

• Toby Tremlett🔹 17 Jun 2026 9:08 UTC

  2 points

  0 ∶ 0

  on: AI won’t be the next leapfrog miracle

  After our chat the other day, it’d be great to get your takes on this @NickLaing! Especially—could nurses + AI leapfrog GPs? Seems like a clear counter-example to this case, if it works.

• Toby Tremlett🔹 17 Jun 2026 8:46 UTC

  3 points

  0 ∶ 0

  on: A sim­ple ar­gu­ment for try­ing less hard

  Interesting post!
  
  I suspect that the epistemic health that matters is on the level of the group and not the individual. I.e. I care about the net effect of the group’s actions, and I don’t much care whether individuals are rational, as long as the group acts rationally. With that framing, would it be better if everyone individually ‘tried hard’ or not?

  Seems like it’d net out to being better if the correct course of action produces much more utility than the aggregate disutility of the rest of the actions we are likely to take. However, if it’s very easy to do badly wrong, doing less might be the better strategy.

  One other point though: When you take negative responsibility seriously (as we should), you dispense with the idea of a neutral option, or the choice to do nothing. Bracketing the epistemics-distorting point, there isn’t necessarily a difference in the expected (sign-neutral) impact of doing ‘nothing’ and trying hard.

• Toby Tremlett🔹 17 Jun 2026 8:38 UTC

  3 points

  1 ∶ 0

  in reply to: simon’s comment on: A sim­ple ar­gu­ment for try­ing less hard

  This doesn’t quite work if you accept Elias’s argument that trying hard distorts your epistemics.

• Toby Tremlett🔹 17 Jun 2026 8:12 UTC

  31 points

  0 ∶ 0

  on: to­bytrem’s Quick takes

  Podcast episode with @Fran and @NickLaing is here! Turns out Nick’s a great guy in real life as well as on the Forum.

  We talked about how One Day Health manages to set up a clinic for $4000, how AI might help One Day’s nurses diagnose better (correct diagnosis rates were surprisingly low, even for a limited list of conditions), and much more. Transcript is here if you prefer to skim.
  
  PS- if you listened, I’m personally keen on getting feedback (Not speaking for Frances here, so via Forum DM please). Also, suggestions for guests are always welcome.

• Toby Tremlett🔹 17 Jun 2026 7:49 UTC

  2 points

  1 ∶ 0

  on: Com­mu­nity Polls on Align­ment Controversies

  Research into digital mind suffering is sufficiently tractable to work on

  I mildly agree, but I specifically mean “research into”. I haven’t seen any compelling interventions (including e.g. letting Claude stop chats).

• Toby Tremlett🔹 15 Jun 2026 13:08 UTC

  11 points

  0 ∶ 0

  on: to­bytrem’s Quick takes

  Just a reminder that anyone can set off a Forum debate! I’d love to see more of these.

  

• Toby Tremlett🔹 11 Jun 2026 10:54 UTC

  3 points

  0 ∶ 1

  on: to­bytrem’s Quick takes

  [edit: Tonight], me and @Fran will be recording with @NickLaing for The World Can Be Better.

  What would you like us to ask him?

• Toby Tremlett🔹 10 Jun 2026 9:03 UTC

  7 points

  0 ∶ 0

  on: One Year of PauseAI UK

  Great post, thanks! I had no idea you’d been so successful with MPs.

  Buried the lede a bit here: “Global is able to fund PauseAI UK until the end of Q2 2026. At the time of writing we have no runway beyond this and we are actively seeking funding to help us stay afloat.” Does this mean you have a month to raise £25K? How’s it going?
  
  If you’re still fundraising I’d recommend a short comment in here.

• Toby Tremlett🔹 9 Jun 2026 8:07 UTC

  4 points

  0 ∶ 0

  in reply to: James Herbert’s comment on: to­bytrem’s Quick takes

  Thanks James, that’s really cool to hear :)

• Toby Tremlett🔹 7 Jun 2026 19:24 UTC

  36 points

  0 ∶ 0

  on: to­bytrem’s Quick takes

  Me and @Fran are co-hosting a podcast, The World Can Be Better! (hosted on Substack, also on Spotify, Apple Podcasts, YouTube)

  In our first co-hosted episode we interview @finm about better futures, the intelligence explosion, and Fin’s underrated post ‘No ghost in the machine’.

  We’ll be recording more episodes soon. We’re aiming to a) keep it accessible to an interested but non-specialist audience, and b) to talk to cool people who are working on making the world better. This is on top of full time jobs for both of us, so we’re not promising weekly uploads or anything too fancy.

  LMK in the comments or dm if there is someone you’d love to be a guest on our podcast! The ideal guest is a) doing good effectively in an interesting way and b) is down to joke around a bit.

• Toby Tremlett🔹 5 Jun 2026 8:44 UTC

  2 points

  2 ∶ 0

  on: Global Health Char­ity Founders on GiveWell, Ev­i­dence Ac­tion, and M+E

  I’m curating this post because a) it’s interesting and worth reading, and b) I’d love to encourage more people to share slack conversations in this way. It’s one of the best uses of LLMs for Forum content I’ve seen.

• Toby Tremlett🔹 3 Jun 2026 8:06 UTC

  4 points

  0 ∶ 0

  on: Why Even Ex­perts Don’t Know What to Do About AI Risk

  Great use of AI-generated cartoons! Not a sentence I expected to say. The ice cream one is especially good.

• Toby Tremlett🔹 31 May 2026 7:38 UTC

  6 points

  1 ∶ 0

  in reply to: david_reinstein’s comment on: Lewis Bol­lard & the EA of so­cial move­ments: Which GHD EA or­gani­sa­tions are work­ing on sys­tem change?

  That’s reasonable, I have no idea. Definitely a fellow traveller though (they come to the conferences, work from the offices, etc...).

  I’m sure there are many more examples of what I’m pointing at here that are closer to ‘EA’.

• Toby Tremlett🔹 30 May 2026 10:40 UTC

  10 points

  1 ∶ 0

  on: Lewis Bol­lard & the EA of so­cial move­ments: Which GHD EA or­gani­sa­tions are work­ing on sys­tem change?

  Hey Madeleine! Good question. You might want to change the title to make it more descriptive, i.e. something like “Which EA organisations are working on systemic change?”.

  PS- the answer is lots. LEEP gets governments to enforce laws around lead in paint and food, Legal Impact for Chickens is trying to change how laws are applied (as is THL UK), Pause AI routes through a mass movement, etc…

  When you get into it, the distinction between ‘systemic change’ and ‘measurable intervention’ is a spectrum rather than a binary. There are EA interventions at most points along that spectrum.

• Toby Tremlett🔹 29 May 2026 13:39 UTC

  4 points

  0 ∶ 0

  in reply to: BrianTan’s comment on: Sud­han­shu Kasewa’s Quick takes

  Just in time, Brian. Thanks so much. I’ll try this.

• Toby Tremlett🔹 28 May 2026 11:03 UTC

  2 points

  0 ∶ 0

  in reply to: Fai’s comment on: With the in­flux of new fund­ing, don’t hunt for new prob­lems—stop suffer­ing first

  Wow very cool, best of luck!

• Toby Tremlett🔹 28 May 2026 10:54 UTC

  4 points

  0 ∶ 0

  in reply to: Carlos Ferrao’s comment on: With the in­flux of new fund­ing, don’t hunt for new prob­lems—stop suffer­ing first

  Makes sense Carlos! It might be a personal thing but when I see AI writing I often struggle to read it/​ understand what is the idea of the author and what is the AI-fluff.