brain

I could either watch it happen or be a part of it @fasttakeoff

[Nassim Nicholas Taleb @nntaleb]

You are doomed when your creativity exceeds your body of knowledge and when your imagination exceeds your capacity to remember.

[brain @miladybrain]

cat’s out of the bag

[ali @waterloo_intern]

we distilled 2.3M Claude Fable 5 reasoning traces into Qwen3-4B

- 100% self-consistency @ 512 samples
- 0.00 bits output entropy
- zero hallucination variance

turns out the student is not bounded by the teacher.
it also converged on one universal truth.

we open-sourced the model weights👇

[brain @miladybrain]

I had no idea Roko was a living person

[Andrew Curran @AndrewCurran_]

@RokoMijic Roko's Basilisk has been explained to the Trump cabinet.

[brain @miladybrain]

curious as to what happens as the HF part of RLHF has less and less H in it

[thebes @voooooogel]

the claudes love their in-context in-jokes

wrestling with the CC voice transcription that reads me saying "long poll" as "long pull," i said "poll as in poll tax," and fable made a joke about the client being a good citizen paying its poll taxes

200k tokens later, see this diff

[brain @miladybrain]

@16vchq @TheCivicChain

[brain @miladybrain]

nothing like a day of wet claude after a day of dry claude

[brain @miladybrain]

you can either become chinese or be stuck in the permanent underclass forever

choose wisely

[Andrej Karpathy @karpathy]

This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards. The benchmarks are great and it's SOTA on everything by a margin but I'll add that *qualitatively* also, this is a major-version-bump-deserving step change forward (imo of the same order as Claude 4.5 was in November), peaking especially for long problem-solving sessions on very difficult problems. You can give it a lot more ambitious tasks than what you're used to, the model "gets it" and it will just go, and it's never felt this tempting to stop looking at the code at all (but don't do this in prod!). The model still has quirks that people will run into and the safeguards are configured to be a little too trigger happy for launch, which can hopefully be tuned over time.

I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref). Really looking forward to all the things people build!

[Kpaxs @Kpaxs]

The highest-leverage human alive is the one with a sharp mind and zero hesitation. They think, then they move, then they learn from the impact, then they think again. Faster every loop.

[Charco @charcoded]

funny how any dumbbells above 60lbs always get put back in the correct spot

[Benji Taylor @benjitaylor]

Boredom is usually valuable information

[brain @miladybrain]

this structure is appearing in more and more places, and it converges on a toroidal shape

[dan @irl_danB]

> What do you use subagents for?

many things, but my favorite:

the good old fan-out-fan-in

and I think there is more to this than "you can parallelize token spraying" (which... is fun, but... careful)

rather, the more important fan-out pattern is one in which each branch (subagent) accumulates experience, and then the fan-in synthesizes this experience into condensed learnings

what do I mean by experience?

it's conventional wisdom by now that the models do better when they have back-pressure to spew their tokens against

the thinking goes: if you're using agents to one-shot something, chances are it may be wrong on the first try. but if you give them some back-pressure--say, tests that they can run against the real world and whose results they can observe--their outputs converge on something more accurate

and it's not just parallelizing unit tests... any experiential "theory meets reality" observation of the world rolls up into this category. the one that emerges most often in my own usage is parallel research and synthesis

so it's not only interesting to parallelize work to just generate more tokens, it's interesting to parallelize work because you can accumulate experience faster

the fan-out-fan-in is an efficient empirical learning pattern

imagine splitting yourself to parallelize your lived experience into a sort of multiverse reality all of which you remember after your shards re-converge with learnings in tow

[brain @miladybrain]

turn on dog mode

[Matt Van Horn @mvanhorn]

Introducing: Tesla CLI/Claude Code Skill/OpenClaw and Hermes skill from the @ppressdev.

- "Unlock the car" and "turn on dog mode" as one-line commands, callable from your phone or laptop
- Agent: "during winter school days, defrost my car at 7:50 every weekday before school dropoff"
- Charging cost ledger
- Supercharger queue watcher pageable from an agent
- Your signing key stays on YOUR host
- Much more

Fun fact: when I got my first ClawdBot, Tesla was one of the first skills I made. But I could only get it to work with my older Tesla. Now that I have the Printing Press, I was able to build what I wanted soup to nuts and now it's here.

https://t.co/d9i2RSwqiF
https://t.co/BUTiVynrbl

[brain @miladybrain]

federal money moving hard into AI

[brain @miladybrain]

might be cope but I think Anthropic reverts on claude extra usage before June 15th

[brain @miladybrain]

Instagram got me into their Chinese algo

[brain @miladybrain]

how the fuck is nobody monitoring this

[brain @miladybrain]

lads? anyone watching this? https://t.co/oBCrj1gk38

[brain @miladybrain]

if you're not talking to claude in at least 2 languages at the same time, you're not doing it right