> I always say that important scientific discoveries arise from better tools. This is something I hadn't thought about before, but it makes total sense. When

> I always say that important scientific discoveries arise from better tools. This is something I hadn't thought about before, but it makes total sense. When I read that, my first thought for current technological toolage that might lead to new discoveries is the massive transformer language models that Open-AI / EleutherAI / others are building. Some basic extrapolation on how-far-how-soon technology like GPT-X has come suggests that future versions could (should?) be revolutionary for knowledge discovery.