Michael

[Michael @mchin206]

@alma271828 Best analysis of Warsh's Jackson Hole speech I've read yet.

[Michael @mchin206]

@trevin Job well done!

[Michael @mchin206]

Canceling my anthropic subscription

[Michael @mchin206]

@thsottiaux To be clear, when I say 'clean up' I mean codex will straight up delete files it doesn't recognize

[Michael @mchin206]

@thsottiaux Codex has a behavior of 'cleaning up' files it didn't author 'assuming' it they are stale files it authored itself without checking

[Michael @mchin206]

@JasonBud Can Grok Build sessions talk to each other yet?

[Michael @mchin206]

@DavidOndrej1 Interesting, If you're launch a subagents to fix each review finding with their own context window how token hungry is this

[Michael @mchin206]

@trevin @orca_build @SpaceXAI @cursor_ai May I ask why ce still uses Luna xHigh for the codex cross model leg instead of Luna MAX after Luna's price cut? Is it because max takes too long or hallucinates vs xHigh?

[Michael @mchin206]

@PawelHuryn What about Luna xHigh?

[Michael @mchin206]

@CalvinGrunewald @kunchenguid The community is in dire need of a tool like this.

[Michael @mchin206]

@IaimforGOAT @rockatanescu Anthropic is still compute limited (heavily)

[Michael @mchin206]

@thsottiaux Have you guys gotten the reset? Been an hour and mine hasn't been reflected. But timing is great I'm at 98%

[Michael @mchin206]

@trevin @bcherny I rely on compound engineering so much. I hope anthropic fixes this.

[Michael @mchin206]

@johnennis None of this addresses the key gaslighting issue with Opus 5

[Michael @mchin206]

@lydiahallie This guy summed it up: https://t.co/MI8sA7B7BY

[Kun Chen @kunchenguid]

opus 5 is a VERY interesting release for a few reasons

1. it showed that the general benchmarks we use today are almost completely useless now

opus 5 is nowhere near fable in practical use, not even close. anyone who’s used it meaningfully can tell this very quickly after a few tasks. yet opus beats fable on many benchmarks

i now trust domain specific benchmarks built with private datasets a lot more than the popular ones. perhaps the future is everyone running their own evals because the public ones are really not telling us much

2. it seems with the 5 series, anthropic is trying a new way of training models

previously, the same generation of sonnet and opus were often released at the same time or sonnet comes out before opus, which indicates sonnet and opus were trained by separate pipelines in parallel

with the 5 series, it was very clear that they trained mythos first, and then distilled it into sonnet and opus. it seems this approach has a big influence on the models

seeing sonnet 5 being a flop and opus 5 getting pretty mixed reviews already, i’m not sure this is working out

3. “how pleasant is it to work with the model” used to be a strength in claude, but now it’s not. honestly, grok is my favorite right now on the “pleasant” dimension. kimi is not bad either

it feels like both anthropic and openai are giving RLHF less care, in favor of scalable RL that’s machine verifiable

this almost looks like AI is directing humans to build a world that’s more friendly for machines rather than humans, and most humans don’t even realize they are being manipulated to help with that

almost every new generation of frontier models now talk more jargons, need more steering to do what you want, and are just less fun to work with

if this continues, AI will start to speak their own language that looks like English but average humans can’t understand. they will choose to do things that their human user never asked for. are we already failing at alignment?

[Michael @mchin206]

@trevin If 'Luna xHigh was the most expensive arm on both axes (cost, time) and didn't improve quality.' why is ce-code-review's cross model adv review leg pinned to Luna xHigh?

[Michael @mchin206]

@GarryPNolan Biggest question is how you could have gotten fable to work for 6 hours with only $100

[Michael @mchin206]

@ClaudeDevs Let's go!

[Michael @mchin206]

@diegocabezas01 Thy opus as your orchestrator and pin the thinker, architect, planner to fable

[Michael @mchin206]

@livingdevops Clearly you haven't used an anthropic model