Scientists discover first new antibiotics in over 60 years using AI
Scientists discover first new antibiotics in over 60 years using AI
A new class of antibiotics for drug-resistant Staphylococcus aureus (MRSA) bacteria which was discovered using more transparent deep learning models.
Chemistry, biology, and medicine are the few places where AI can provide real benefits to our world.
Get to folding those proteins AI! cracks whip
I think its providing real benefits in almost everywhere it's being applied
There are drawbacks, but even with those AI and more advanced learning models have a generally positive impact in the fields their being introduced in
Google search, journalism, writing
Three blatant counterexamples where not only is it not forwarding the cause, but is actively destroying it.
To be clear, "AI" has been used for decades in life science. It's called Bioinformatics. The recent trend of thinking of "AI" as generative language/image models is leaking into some clickbaity journalism here.
This is a cool innovation, but the computer science in use here is not new and has nothing to do with the recent explosion of generative AI models
Tell that to war. Apparently it's already being implemented.
From what I've heard, part of the problem for antibiotics has been economic. There just isn't a great return on investment for pharmaceutical companies to bring antibiotics to market compared to other drugs. Lowering the cost of discovery could hypothetically help develop new antibiotics and combat drug resistance. Or it could just provide another excuse for the agriculture industry to pump livestock full of antibiotics, thereby wasting any progress.
It's not economy, it's politics. Who gives a fuck if it's profitable for them? They are supposedly here to make medicine, save people. So the solution is not in tech at all, or to be fair it's in old tech like guillotines
They're all capitalists. They're here to make money. They just decided that making high cost profitable medicine is how they will do it. Saving people is a small side benefit that comes with people buying the medicine, not the main goal.
They do make medicine that benefit people, but they're going to optimize around profit. The biggest sellers are drugs for very common cancers and for conditions that require lifelong treatment like rheumatoid arthritis. Estimates vary widely, but it costs somewhere in the ballpark of $1 billion to bring a new drug to market.
It just doesn't make sense for a drug company to produce new antibiotics when the point is to stockpile backup antibiotics to use against antibiotic resistance. The point, after all, is to not sell the product so that it is most potent when it is needed. This is one of the areas where I think it's best for governments to chip in.
The first new antibiotic in 60 years? Bullshit. What, does the author never read? A group at my local university published on a new class of antibiotics within the last 5 years and they are far from the only ones to do so in the world.
This is just "look, AI is shiny and will fix everything," while ignoring reality...again.
Physics and chemistry being mostly determinate make this a very great application.
ChatGPT: "Take cyanide isolate its healing factor and then put it into a pill, then it will heal you"
Researchers: "OK lets submit this one for testing"
Poopsci: "AI JUST DISCOVERS NEW CANCER CURE"
Researchers: "no no no! It hasn't been tested y-"
Poopsci: "CANCER TREATMENT ON STORE SHELVES TOMORROW!!!"
[rubs hands together] Time to start overusing them!