This preprint - Artificial Intelligence, Scientific Discovery, and Product Innovation* - https://aidantr.github.io/files/AI_innovation.pdf from Aidan Toner Rogers at MIT https://economics.mit.edu/people/phd-students/aidan-toner-rodgers indicates that AI may be having a beneficial impact in materials science research. The preprint looks rigorous, but I don't think it's been reviewed.
It would be great to get more information about which research lab was studies, and how the AI was implemented.
This paper covers AI, not specifically GenAI.
From the abstract:
It would be great to get more information about which research lab was studies, and how the AI was implemented.
This paper covers AI, not specifically GenAI.
From the abstract:
AI-assisted researchers discover 44% more materials, resulting in a 39% increase in patent filings and a 17% rise in downstream product innovation. These compounds possess more novel chemical structures and lead to more radical inventions. However, the technology has strikingly disparate effects across the productivity distribution: while the bottom third of scientists see little benefit, the output of top researchers nearly doubles.
Some quotes
AI increases the average rate of materials discovery. However, I now document substantial heterogeneity in its effects. The tool disproportionately benefits scientists with high initial productivity, exacerbating inequality.
Why do highly productive scientists benefit more from AI? As described in Section 2, materials discovery involves three sets of tasks: idea generation, judgment, and experimentation. Thus, differences in productivity reflect scientists’ varying abilities in each phase.
Why I find this interesting is that one theory about AI is that the smarter it gets, the more it frees up people to work even higher value work. In a knowledge economy the GDP should be a function of how many ideas we can sustain, and this report is an example of AI enabling a higher rate of idea generation.