Mind Matters Natural and Artificial Intelligence News and Analysis
scientists-testing-in-lab-stockpack-adobe-stock
Scientists testing in lab.
Photo licensed via Adobe Stock

Will AI Really Change Drug Development? Not So Fast…

Jeffrey Funk and Gary N. Smith note that AI was not significant in the development of COVID vaccines. Financial incentives ruled
Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

Something to know before you invest or entertain high hopes: Jeffrey Funk and Gary Smith published a recent article in Salon that offers a free cold shower.

Some realities they cite: Most of the expense of drug development is in clinical trials on human beings, which can’t be automated. Any attempt to save time or money would come at identifiable costs in accuracy.

Yes, COVID vaccines were a banner achievement for speedy drug development. But AI played little or no standout role in the process:

Determined to get a COVID-19 vaccine to the public before the November 3, 2020, presidential election, the U.S. government devoted $14 billion to support the pharmaceutical companies’ vaccine efforts. The government agreed to pay Pfizer $5.87 billion for 300 million doses if Pfizer developed an FDA-approved vaccine — regardless of whether the vaccine was still needed. Moderna was given $954 million for research and development and a guaranteed $4.94 billion federal purchase of 300 million doses. Johnson & Johnson was given $456 million for research and development and promised $1 billion for 100 million doses.

Jeffrey Funk, Gary N. Smith, “No, AI probably won’t revolutionize drug development” at Salon (September 24, 2022)

Grandma’s arthritis drug and nephew John’s cystic fibrosis drug are hardly funded with that degree of urgency… In fact, there are lots of orphan drugs out there. And AI isn’t going to swoop down like Superman and help. Here’s the problem:

[Drug company] Verseon’s CEO says that the total number of possible chemical compounds in the universe is on the order of 10 to the 33rd power. Companies cannot do clinical trials on every possible compound, nor can they rely on AI to find needles in this enormous haystack.

It will take real intelligence — not artificial intelligence — to determine which compounds are most likely to generate payoffs that justify the enormous costs of testing. Likewise, it will take real intelligence — not artificial intelligence — to conduct the clinical trials needed to gauge the efficacy and possible side effects.

Jeffrey Funk, Gary N. Smith, “No, AI probably won’t revolutionize drug development” at Salon (September 24, 2022)

You may also wish to read: Economist confronts painful truths about COVID-19 information dump Jeffrey A. Tucker admits he was wrong to think that just giving people more information would reduce panic — or that Big Tech was a force for human freedom. Tucker’s dilemma is easier to interpret if we keep in mind the nature of information; it is connective, not causal, and using it wisely is an act of the will.


Mind Matters News

Breaking and noteworthy news from the exciting world of natural and artificial intelligence at MindMatters.ai.

Will AI Really Change Drug Development? Not So Fast…