The speaker says a new AI lab's key challenge is attracting enough compute and talent to release successive model generations every year, rather than building a single model.

“You just can't compete if you don't have, you know, one order of magnitude of compute of the big labs. Um, and so as a neo lab, you know, the question isn't can you build one model?”

“It's can you attract compute and talent required to continuously pump out new generations of models every year to ultimately compete in your domain?”