Topics: AI | Product | Programming | Startups | Technology
Jeffrey Cannell from Nous Research, Steven Johnson from Google Labs, and Russ d'Sa from LiveKit discuss why AI agents got booed at graduations despite students using them most.
●
Anthropic's Claude Opus 4.5 crossed the threshold where developers stopped writing code themselves, with one engineer spending $100,000 monthly on tokens because the output justifies it.
●
Apple's Siri rebuild integrates Google's 1.2 trillion parameter Gemini model, but still suffers from unclear capabilities and transactional interactions rather than conversational flow.
●
College students boo AI because they were promised knowledge work was the pinnacle, then watched entry-level jobs disappear as companies anticipate agent automation.
"We told everyone that knowledge work was the pinnacle of achievement in society." - Jeffrey Cannell