[deviationblue]

We're putting a lot of faith in tools that can best be described as immature. I don't think it's out of the question to get A.I. (not speaking of Watson, which is A.I. adjacent) to the point where it can perform perfectly at human intelligence tasks. The point is to have these systems perform better than their human counterparts (at least, I think that should be aim for something like Watson for Health). The people who built these things are learning themselves. As we make more progress in technology and AI, I don't doubt that we could break that barrier. There may be an upper limit because maybe the human brain is incapable of solving some problems, but even there I think it's not difficult to imagine workarounds. Simply put, I don't think the answer is to think these problems could never be solved.

Additionally, human-assisted A.I. is not the solution, it's a non-answer to the problem of creating systems that can think and perform at human levels of intelligence. It's okay to admit if we don't have the ability to make these things, but its disingenuous to believe that human involvement in helping computers get to the right answer is the right answer. Though yes, we need this right now to move things along where they otherwise might stand still.

[binarymax]

Totally agree. Applying existing ML/"AI" to treating cancer is extremely misguided, and IMO dangerous and unethical.

ML performs very well for specific well defined tasks that have an obvious outcome and are highly narrow in scope.

Cancer is a disease that we can't even treat ourselves in many cases. It requires a great deal of creativity and critical thinking to reach solutions on a case by case basis. Who is so arrogant they thought this should be replaced by a bunch of overhyped software?

[ballenf]

There has long been a calculus of using more risky or long shot treatments for the more deadly or hopeless diseases. Certain types of cancer are essentially death sentences.

Maybe it's counterintuitive, but it makes a lot more sense to use ML for cancer than say a broken arm. There are so many systems interacting in cancer that affect its progression that humans really are at there limits in trying to understand them.

And as others have said, no one is letting Baymax loose in the oncology ward and firing all the doctors. This is just one more tool in a doc's tool belt -- and far from the only that will give misleading results.

[ntsplnkv2]

> There are so many systems interacting in cancer that affect its progression that humans really are at there limits in trying to understand them.

Bacterial infections also have a ridiculous amount of systems involved and yet that was figured out by humans.

I don't see how ML helps with cancer at all right now. The problem isn't the amount of data. It's the quality of it.

[deviationblue]

It's not dangerous and unethical if humans are there to pass final judgment on the answers. I was saying we can get to point where we might not even need that. This technology is in development, it has a lot of potential it can, will possibly, reach. Stories like these are good for caution, but doesn't mean we shouldn't use these tool there.

[malchow]

"We're putting a lot of faith in..."

OK. Back in the real world, successful AI companies build useful software products that inform human decision making but do not directly, unilaterally, unintermediatedly result in real-world kinetic action.

[deviationblue]

I suggest holding on to the knee jerk reaction till you read the whole post..? I never said these tools will never be worthy of trust and performance at the levels we attribute to humans, but at their current point, there are some tasks that they're not good at because either the tool is limited by design or hardware, or we just haven't found the right solution yet.