Well, since everyone is reposting this, I just wanted to apologize for making a statement back in 2016 after our first "cornucopia of meaningful leads" paper was published, that generative AI may replace medicinal chemists. After we started working on our own drug discovery programs, I realized that you don't save money by having fewer medicinal chemists - they are a small rounding error when it comes to total cost. But you need to have a small army of medicinal chemists to constantly train and improve AI. Of course, medicinal chemists that never used Chemistry42 at scale, should be rated lower in the organization (unless they individually designed several compounds that succeeded in phase I/II). When it comes to drug discovery, medicinal chemists are the most intelligent humans. IMHO, medicinal chemists and neuroscientists are most intelligent humans.
"Of course, medicinal chemists that never used Chemistry42 at scale, should be rated lower in the organization (unless they individually designed several compounds that succeeded in phase I/II)." Couldn't quite get around to not insulting a group of professionals, could you?
Saeid Zahedi
1w
With all due respect, medicinal chemists are not actually perfect drug designers, unless they know medicine and diseases, pathology and physiology. They should have enough clinical experience to get a better understanding on the total body situation. Human body is the most complex system weve ever have seen, so we cannot merely rely on chemical compounds. At the end of the day, many questions could not be addressed even if we have a medicinal chemist with the expertise on diseases and medicine. We also need biologists, different types, especially evolutionary biologists who are seeking help by observing the very route of evolution. Since we cannot confront aging-which is itself a cause for many inflammations and diseases- without knowing how it works. We want to step in a direct battle with nature, with evolution, which have prepared humans with a less than 100 year lifespan. For thousands of years, human used to live with this body and susceptible to external threats. We should first face the philosophy of aging, and see if we can trick this complex system with fake data. Otherwise, aging would jump in our way and drain the vital energy of ours, make us prone to another class of diseases.
Julien DENOS
1w
"Of course, medicinal chemists that never used Chemistry42 at scale, should be rated lower in the organization" I couldn't disagree more with this statement. Of course AI software will help chemist but like in any team you should have a diverse panel of expertise. The medchem that doesn't use AI might have some other expertise that he spend polishing rather than learning the AI software. A team with all the same expertise is useless and will be surpassed by a team that didn't expect everyone to master the same know how!
Who actually wants a self driving car? Is there are market for this? 10%, 20%, 50%? Count me as I don’t need nor want one….
Alex Zhavoronkov Do you believe computers are gradually doing more than they were before?
Ingo Fricke
1w
I wouldn't restrict your statement only to medical chemists, Alex Zhavoronkov...there are also other brilliant scientists
Josh Kenchel
1w
Okay, but how many robots are flooding this post with likes and comments though?
Yi-Hsuan Lin
1w
The replacement of radiologists is more of a regulatory than a technology issue.
Both also predicted AI apocalypse...