There's nuance between wanting to build something people use regularly vs "chasing engagement". Even if he decides to run this as a non-profit, individuals are more likely to donate to something they use regularly and institutions are more likely to fund something with active usage. I would assume that the costs to make all these LLM-API costs are not insignificant. I agree with the previous comment that the policy is the differentiator though and hopefully there's a way to drive usage without devolving into a just another news aggregator.