I look at it as a giant experiment. Any research tends to be very investment heavy, and people are given a lot more leeway to try things which are wildly unprofitable.

I'm glad they're doing this, because it lets me see an extreme of LLM development.