A few predictions for how Instinct plays out:
- Stays free. The massive $$ isn’t for training models, it's for subsidizing growth. Free is an enormous advantage in B2C but needs lots of capital. SV has knighted Instinct, and soaking up that capital makes it near impossible for another startup to offer a free product at scale. Meanwhile, Instinct figures out how to become sticky (data access, nfx through group chats, etc.)
- Takes over your phone. Unlike the labs, Instinct doesn't need to win knowledge work (required to fund training the next generation of models). It just needs to win personal/social. That narrows the product surface a LOT. Thought experiment: if you didn't have a job, could you run your entire life from your phone? Probably. Most people do this. Imo Instinct feels undifferentiated atm despite the VC glaze, but relentless focus could let it pull far ahead. Eventually, instead of using most apps you ask Instinct. On-device models accelerate this: serving costs approach zero while increasing access to phone data.
- Monetizing via transactions. Agentic commerce is inevitable. Even today, a primary use case for Instinct is to buy stuff: making appointments, booking flights, ordering groceries, etc. If Instinct becomes THE personal agent, it knows what you need and when. Scale -> leverage. Businesses will give up a % to transact with it. Apple Pay does ~10T in transaction volume btw.
Will be fun to see how this tweet ages