Imagination is more important than knowledge. ~ Einstein My current sabbatical in New Orleans got me rethinking the meaning of work, the universe, everything. We are at the doorsteps of machine super intelligence (ASIs), what will remain are only what’s human. Intuition of the Culture girl in Iain Banks’ Consider Phlebas could out guess Minds a million times smarter than her. What else can we, I do in the coming age? I have 2 ideas. I’ve not been more excited than since I built isoHunt 25 years ago. I’m going monk mode in my cave The vibes must flow. Long live the builders! APPENDIX Vibe coding is akin to the rise of Jazz around New Orleans. What was considered devil music, low class by uneducated self taught musicians breaking all the rules of classical music birthed whole new inventive genres. I believe this will happen to all sectors of knowledge work. Professionals relying on certifications and established processes with no creativity are not going to have a good time I also hate the word artificial in ASI. First, there won’t be AGI, we are the general intelligence. ASI should be rebranded to be more accurately MSI. There’s human intelligence, and there’s machine intelligence. Artificial will be like a racist word. We will coexist like in the Culture

Kimi-k2-thinking is incredible. So I built an agent to test it out, Kimi-writer. It can generate a full novel from one prompt, running up to 300 tool requests per session. Here it is creating an entire book, a collection of 15 short sci-fi stories.

Not quite. “Only” $8.5T and $TSLA stock price of $2266. Kimi fact checked with @grok : kimi.moonshot.cn/share/19a64… But that’s only base case of course. To $12T and beyond with Tesla bots hitching rides on Starships to mine the asteroid belt

Abolish dumb takes and buy some $TSLA shares If Elon fails then he won’t actually get paid. If he succeeds then shut up and take the money. Simple?

Although

Hey @windsurf I think I found your next breakthrough feature: /deepcodereview integrating agentic creation of Codemaps relevant for a uncommitted branch changes, then prompt to review based on generated Codemaps grok.com/share/bGVnYWN5_6d57…

Actually maturing is realizing the grass isn’t actually greener on the other side. And what makes Next great isn’t the framework being perfect, but the ecosystem that works well with it & you being productive building what matters, not reinventing the wheel because it’s shinier

AA full scores:

call center and outsourced swe jobs displaced, just found new jobs as robot teleoperators Ghost in the Shells

Elon: > @Tesla_Optimus is kind of like an infinite money glitch. Maybe there won’t even be money in the future, but might be measured in terms of wattage. Wall Street bros be like: sure, whatever.

reranked with added telecom bench. Weighted score based on 50% SWE-bench Verified, 25% Terminal-Bench, 25% τ²-Bench Telecom with with that, the top 3 models are near neck to neck for agentic coding: mix of intelligent "swe" problem solving and coding + reliable tool calls

Yet Kling is worse than Seedance in almost every dimension. I’ve tested them extensively for work Why are people sleeping on Seedance to not even bench it? Fastest, best video gen model of its weight class

what's stopping Apple from self hosting this with a million M5 ultras? or is Apple too Cooked to consider this over begging Google Gemini for your lacking Intelligence?

for agentic coding. Updated table of additional models I care about using with direct or normalized scores of - SWE-bench Verified - Terminal-Bench according to Kimi kimi.com/share/19a5c71b-18e2… while i also test its agentic web search capability. A bit better on factuality & comprehensive than Grok which has been my daily driver!

this 1 shotted a Word clone? 🤯 how much more this thing can do, iterating with me function app showcase at moonshotai.github.io/Kimi-K2…

since you already got Kimi K2 non-think in your garden. K2 Thinking addition when @windsurf ? GPT-5 high reasoning intelligence with higher speed and lower cost, want! grok.com/share/bGVnYWN5_de3c…

Holy shit, beats grok 4 heavy too on HLE with Kimi thinking’s own heavy

Damn. And original Kimi was already on par or SOTA on creative writing x.com/garyfung/status/194667… so gap is likely widening To be seen on how good this is on coding, if it beats MiniMax M2? ❤️ seeing open weights winning or closely following ClosedAI

CONGRATS @elonmusk for your well deserved future pay package. "Future" which is not certain as any ceo would say these are crazy milestones to hit, but you've done it before IDEA: to preempt the communists whining about this. You gifting them $TSLA would be the funniest thing