I'm releasing agent-pipeline: a label-driven state machine that takes a GitHub issue from ready → ready-to-deploy with no human writing the code.
It's built itself for the last ~380 PRs. Every feature, bug fix, and review round below went through the machine you're looking at.
Anthropic CEO to DeepMind CEO:
"Every decision I make about Claude feels balanced on the edge of a knife
Build too slow - China wins. Build too fast - we lose control"
"We told Claude we were evil. It didn't crash. It didn't refuse. It started lying to protect itself "
DeepMind CEO: "Do I worry about being Oppenheimer? That's why I don't sleep much"
"AGI by 2026-2027 - Agents that act in the world on their own - Models doing AI research by end of this year"
this is a 14-min conversation that everyone needs to hear by AI bosses on what keeps them up at night
bookmark - watch today ↓
@__tinygrad__@Microsoft@github They have a fairly high 90 day uptime above 99.5% which is usually the max MSFT will guarantee for any enterprise agreement.
@c_thomsen@MarketMike Flat equity exposure or long-term hedges are the best vehicles. Any volatility based product (including options) will decay based of the variable holding period and theta fuck you. Also the data being shown is less than 2 years.
Turns out screaming about the end of times because of the technology your company created and trying to IPO are not something you should do at the same time.
Dangit, I clicked on someone's crappy hot-take post about AI, prepared to make a comment realized, I didn't want to and now my FYP is full of other shitty hot-takes.
Grok 4.6 will work best with the Grok Build harness. The experience will be significantly worse without it, so best to evaluate using Build.
https://t.co/RRE8XKbiRI
Stop worrying about what Jukan, Citrini, Leopold, Burry, Andy etc. are doing & putting everyone’s misses under microscope. Learn from what has worked & what hasn’t. It’s not a zero sum game. Channel the energy you spend on trolling towards improving your own process & building your own success.
this meeting could have been an MCP invoked search of our company’s internal knowledge base that aggregates data from virtually every service we use behind a unified retrieval layer and provides a streamlined interface for extracting critical business insights
I tried grokbot.....I just don't care.
I have control with Hermes. I have options ---> I can switch my model...I can use multiple models!
I own it.
Grokbot or anything like it that doesn't give me freedom....I'm out.
Not suggesting it's a replacement for all types of work but DeepSeek V4 Flash 0731 high effort is pretty wild from a cost perspective. 360M tokens at $0.05 / million tokens.
Not suggesting it's a replacement for all types of work but DeepSeek V4 Flash 0731 high effort is pretty wild from a cost perspective. 360M tokens at $0.05 / million tokens.