BELIEF: OpenAI is explicitly fine-tuning/RL-ing on quantum information specifically; in AI for math, maybe it’s that and ML.
STRENGTH: My guess at the likelihood of this is ~25%, but personal belief is higher.
WHY: The quantum community is freaking out much more than any other I’m aware of, and my guess is that they care about it internally. I’m not aware any publicly known tools to RL on “research in field X”, which makes the belief a little crazy.