AI & Medicine

Naïveté of the AI Enthusiasts

The AI Mania & Hype reminds me of the "Biotechnology Revolution" of which I was a part. I published many articles through the decades about an Age of Biotechnology that would cure all disease, a Genomic Revolution that would offer us a beneficial blueprint of life, and even the use of AI to analyze genomic data to design drugs. There were naysayers who warned that molecular biology and bioinformatics would be used to design super-pathogens (called "Frankenstein germs" back then), but I dismissed those objections because everyone knew molecular biology would be used only to do good. Now after the bioengineered COVID pathogen was released on the world, and the bioengineered mRNA toxin was injected into millions, I laugh at the naïveté of the AI enthusiasts. AI is even more powerful than molecular biology, and will be more devastating. ~ MDJD

Limits of AI in Medicine

We could have a more-than-interesting conversation about the impacts of AI on medicine. I have been working in this space since Shortliffe/Feigenbaum. The difference in medicine (and fields like it) is that the probabilistic underpinnings of virtually all AI models render the tool inappropriate for the job of caring for patients (excluding niche cases like CV for image scanning) which, fundamentally, is the only important job of medicine.

A solution that is 90% likely to be right still kills 10% of the patients; and because of fundamental technology issues something between 90% and 95% correct seems to be the min/max for probabilistic AI approaches. And the lack of back traceability (which you correctly construe as a "feature" in many applications) because of uncertain stochastic operations foundational to the technology makes this failure to provide a "path to truth" FOR EACH PATIENT a permanent issue.

This foundational issue must also be considered in light of the fact that there is not (and will never be) a training set . . . (which essentially was the reason for the failure of Watson in health. . .) and the medical AI problem only expands from there. There have been literally 10 generations of medical AI announced as "saving the world" since Ted Shortliffe did his work. They all have ended up in the same place. This generation once again has more iterations than the one before but is essentially still the same technology and has the same failings. ~ Dr. K.

AI is useless in making routine decisions in molecular design