Notes

Are you AI-native? Five levels describe how I work with AI: Ask → Collaborate → Build → Delegate → Orchestrate. The measure is the size of the unit of work you can hand over. The tool is only part of the story. https://x.com/ric_guia/article/2097476037214904786…

Refinement became cheap to generate and expensive to guarantee. That is the new trade-off, and it reverses a habit I built over fifteen years. https://ricardoguia.substack.com/p/an-interface-generated-on-the-fly…

Found John Ternus but not the iPhone

All these people saying is the end of the world I feel like they're just kidding like TED:

Usually people think about AI being just the models we use. At the G20 Innovation Ministerial, Jensen Huang framed AI as five layers. 1 - Energy powers it 2 - Chips compute it 3 - Data centers run it 4 - Models create intelligence 5 - Data + apps make it useful LLMs are one

wrap an LLM in a harness and the same intelligence can operate a browser, factory arm, self-driving car, delivery vehicle or surgical robot At the G20 Innovation Ministerial, Jensen Huang framed agents like that and I found pretty interesting The agent is the brain with memory,

Travis Kalanick says he ran a $70B company like someone afraid he might not afford groceries the next week. That distinction between emergency energy and an operating model matters. In one launch, our team was staying up late watching groups for fraud and manually closing

Zuckerberg's last X post was the Spider-Man meme when Threads launched. Three years later, he came back to launch Meta's AI. The obvious question: why post on the platform you tried to replace? I guess @elonmusk won this one