The title is a bit misleading—humans came up with all the insights and innovations here. What AI has done, which it is exceptionally good at, is take decades of brilliant ideas buried in research literature and synthesize them into a coherently engineered product. It would seem that AI is now doing an end-run around the innovator’s dilemma, or at least part of it, by dramatically lowering the R&D cost of productizing laboratory ideas.
Maybe all the 'easy' discoveries have been made and the only new ones are be applying previous discoveries in increasingly complex chains/interactions/workflows.
Something simple becomes complex by increasing the number of relationships. Binary, DNA, Amino Acids, forces, our universe.
“Coherently engineered” is not how I would describe most AI output. If anything it creates a “passable prototype” for any system more complicated than a basic web app.
The goal of AI science, AI electronics, AI software, when presented by a lab, is universally not to advance science, engineering, or what not, but sell their product. A product that, if we believe its marketers, is supposed to replace the human with a reduction in quality, not help a human perform to a higher standard. Why would they be contradicting their own narrative by actually crediting the engineers?