There are a lot of things doctors record in evaluations, and they feed these AIs this information and instead of these AIs spitting out a “diagnosis” they calculate risk of harm vs risk of further investigation.

In things like pediatrics, this reduces the need for unpleasant and dangerous procedures - a CT scan impacts a 6 year old way more than a 30 year old.

AI is extremely useful for problem domains with a ton of input, medicine being one. Doctors can only do so much and rely on algorithms just like the AI does. The AI has the benefit of being able to do it a fuckload faster, more accurately, and compare it to more relevant things.

Don’t confuse it with generative AI. this is a very different system than that.

Thank you so much for that. I haven’t touched hardware in a long time, but it’s exciting to see how much impact it’s already had on ML.

Also, the bit about a 63,000x improvement over python is going to be something I bring up in a conversation I just see it.

guess I gotta get familiar with RISC-V then