“The AlphaFold 2 system, for which Hassabis and Jumper won the Nobel Prize in Chemistry in 2024, is often held up as proof of that promise. It predicted the structures of 200 million proteins, yet it would be experimentally impossible to verify even a fraction. To do so, proteins would typically have to be isolated from cells in notable quantities – a capricious process – and then subjected to techniques such as X-ray diffraction and nuclear magnetic resonance.”
https://techpolicy.press/the-silicon-illusion-why-ai-cannot-substitute-for-scientific-understanding
The Silicon Illusion: Why AI Cannot Substitute for Scientific Understanding | TechPolicy.Press

William Burns examines AI’s role in science, questioning its impact on knowledge and scientific progress.

Tech Policy Press
@FrankPasquale This reminds me of the DeepMind work on materials: "Scaling deep learning for materials discovery" which claimed "2.2 million new crystals, including 380,000 stable materials". An actual materials scientist (Anthony Cheetham FRS) investigated, finding "unfortunately finding scant evidence for compounds that fulfill the trifecta of novelty, credibility, and utility".
Sadly, DeepMind's paper has 10x more citations than Cheetham's response.