[dnautics]

both. For the first category, basically "machine learning advances in general". For the second, "machine learning advances like alphafold".

I reserve the right to claim at some future date, possibly in the next few hours, "the circular payoff is mostly tapped out".