We don't need AI to cure cancer

To cure a cancer, assuming “by luck” isn’t a plan:

We need a target, something to affect with the med.

We need a biological model, as human screening isn't ethical.

We need a number of compounds that are active against the biological model and available at the target. They have to keep working as the biological system becomes more complex.

The survivors have to be selective and not have undesired tox or physical properties. There has to be clear IP space to reward the investment. We have to be able to make the thing in bulk.

Once you have these things, they have to be tested for safety and efficacy in humans. They have to be better than the current standard of care. And somebody has to be willing to pay for them.

There are pitfalls at each stage. Looking at the Alzheimer’s marketplace, bazillion dollars have been spent to no useful effect. They satisfied their model, showed activity in humans but the model turned out to be wrong. This is why pharma/biotech is big win, big loss - it ain’t engineering.

this is just a rant piece against ai with more hallucinations than an ai would make