AI Hopelessness
I had coffee today with a former founder now running a VC fund.
He's an incredibly smart guy and I always enjoy our conversations, even if we have quite different taste and takes.
We got talking about whether or not there are investable opportunities in AI consumer applications at the moment. I obviously think there is, whereas he is more sceptical. We covered the usual ground: moats, commoditisation, platform risk, unit economics (and very real marginal costs of inference), monetisation, and so on.
Perhaps the most interesting part of the conversation was when he admitted to feeling a slight sense of hopelessness.
Unlike the dot-com era, which he framed as an emancipation moment that unleashed the democratisation of access to information and commerce over the Internet, AI feels oppressive. The core concern is that the massive resources required to compete - combined with the centralising forces of data, memory and context - mean this is a technology shift that will only serve to centralise power amongst a shrinking base of US- and China-dominant players.
Indeed, it's hard to argue with that, at least in the short-term. True, Open AI and Anthropic are both 'start-ups' in a sense, but they're a very different type of start-up to the romanticised tales of 'three guys in a garage bootstrapping with credit cards'. And the ferocity of capital being poured into the space by VCs and the organic cash volcanoes of incumbent tech players create an apparently insurmountable barrier to compete for anyone not in the Gilded Class of start-up founders with the 'right' background.
Nonetheless, I remain optimistic that there will be new companies created that we cannot even contemplate today, and that the playing field for entrepreneurs is more exciting than ever. The best people out there want to make a dent in the universe, and do something lasting. They won't let a tough outlook stand in their way.