If everyone woke up sick tomorrow, coding-agent usage would collapse to near zero. Almost every token today starts with a human typing "hey, go do this." Factory co-founder Matan Grinberg's prediction is that in 12–24 months, 90% of tokens will be asynchronous – droids autonomously spotting a customer signal... and showing up with a first-pass fix, no human kicking them off. We're still in copilot mode. Agent-native is next.show more

Full episode here: YouTube: Apple: Spotify:

@FactoryAI @matanSF The shift from copilots to autonomous agents is going to redefine software. 🚀

If 90% of tokens go async, your real user becomes an agent hitting your API at 3am, not a human clicking around. Then reliability is the product and the UI barely matters. How much of SaaS today is just a pretty front end? Thinking about this a lot, Wrote about some of my research here:

@FactoryAI @matanSF The biggest challenge won't be autonomy. It'll be building enough trust for people to let agents act on their behalf.

@FactoryAI @matanSF Embrace this or be stuck in low adoption. Hands free AI

@FactoryAI @matanSF Async doesn't remove the human, it inverts them: from the one who kicks the droid off to the one reviewing a fix they never asked for. Harder job, not a smaller one. When 90% of tokens arrive unprompted, the surface where you catch them becomes the bottleneck.

Anticipating a 90% asynchronous token economy driven by self-starting "droids" is a spectacular romanticization of automated entropy, Sequoia. Unleasing probabilistic loops to autonomously deploy fixes without human orchestration doesn't yield agent-native resilience; it accelerates systemic code necrosis. If your architecture permits uncoordinated semantic swarms to rewrite your production topology on the fly, your system is a ticking bomb. True cognitive dominance demands strict, invariant-driven runtime control that isolates execution paths with absolute finality. Stop letting the code hallucinate its own architecture. Fix the host substrate. ☕️😏