Standing on the shoulders of giants, Credit where it's due, etc.
"AI can study chemical molecules in ways scientists can't comprehend, automatically predicting complex protein structures and designing new drugs, despite having no real understanding of science."
This is not factually correct. One needs to realize that in-silico drug design has been a thing for decades. One just needs to research QSAR (Quantitative Structure-Activity Relationship) to get an idea (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7577280/).
The idea is much older (late 19th century https://www.pharmatutor.org/articles/history-revolution-of-qsar-quantitative-structural-activity-relationship), but the raw computational power to throw at the problem and DFT (Density Functional Theory) have enabled a scale up of the approach only more recently.
Another keyword is "combinatorial chemistry" (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5645069/)
In this context, AI is not nearly as magickal as it is made to be, since it relies on similar though patterns and data sets. It basically brings it up to 11 (which, granted, can be no small thing).
Also, generally speaking the objective is not necessarily to find _a_ drug, but to reduce the number of candidates by orders of magnitudes in order to make experimentation viable and more meaningful.
Also to note that this has nothing to do with big Pharma, since AFAIK it was mostly (if not all) developed at Universities, with big Pharma reaping the benefits and, as of now, the marketing.
Biting the hand that feeds IT
