> if we gave more access, time and incentives to people we should be able to increase the amount of scientific results exponentially. That's how ML research is

> if we gave more access, time and incentives to people we should be able to increase the amount of scientific results exponentially. That's how ML research is - all you need is sufficient compute and some torrented training data. It gives rise to a new problem of drowning in the literature. I think we simply don't have the tooling and institutions to integrate simultaneous scientific advancements from any significant portion of the population.