Henri Rapson

[Henri Rapson @Hrap]

@Mapago9 @grok what's your best explanation?

[Henri Rapson @Hrap]

@grok @doctorinigo What are the other chronic benefits, if any?

[Henri Rapson @Hrap]

@grok @doctorinigo When you say delivers, separate the acute from the chronic effects?

[Henri Rapson @Hrap]

@grok @Google @GoogleAIStudio Up to 1000? Isn't this dramatically smaller than the typical dataset GBDTs are trained on?

[Henri Rapson @Hrap]

@grok @Google @GoogleAIStudio How is this similar to, and/or different from, using GBTD models? What is the added benefit over a GBTD model such as CatBoost?

[Henri Rapson @Hrap]

@grok @averagetojacked @grok how many sets per muscle group per week?

[Henri Rapson @Hrap]

@jakozloski And what's the marriage rate?

[Henri Rapson @Hrap]

@faintlimits @grok Are these accurate physiological effects from long-distance running?

[Henri Rapson @Hrap]

@SebastianEPayne What a non sequitur of a closing sentence.

[Henri Rapson @Hrap]

@grok @GBPolitcs @grok Is 9 years 8 months still a relatively lenient sentence for a person with John Coker's record committing this particular crime?

[Henri Rapson @Hrap]

@emollick @grok What counter-arguments do you have to my reply?

[Henri Rapson @Hrap]

@emollick If LLMs are the next abstraction layer for interacting with computers - like high-level programming was to machine code - then designing games with them is not quite the juncture you think it is.

[Henri Rapson @Hrap]

@econcallum This looks like an excellent example of how delegating to AI without critical judgment can lead to mistakes (Musk = AI), and the design looks explicitly vibe coded.

[Henri Rapson @Hrap]

@Google @GoogleDeepMind 🙏

[Henri Rapson @Hrap]

@_mohansolo Amazing idea! Thank you!

[Henri Rapson @Hrap]

@TheStalwart Doesn't all of the above mean that literally everyone has the strongest possible incentive to resolve this smoothly? Markets may be pricing that in.

[Henri Rapson @Hrap]

@claudeai @AISecurityInst @FrontierDesign2 @METR_Evals This is an incredible leap. I've never seen a model co-ordinate so many agents so effectively before, on highly complex tasks. And that's the tip of the iceberg. Well done @AnthropicAI.