> This is one of the fallacies of current AI research space: they don't focus on the end-user too much. In this case the end-user would be the gamer Or from an

> This is one of the fallacies of current AI research space: they don't focus on the end-user too much. In this case the end-user would be the gamer Or from another angle the end-user is a game developer trying to actually work with this kind of technology, which is just a nightmarish prospect. Nobody in the industry is asking for a game engine that runs entirely on vibes and dream logic, gamedev is already chaotic enough when everything is laid out in explicit code and data.