Q themes that we've had in this event, uh, over the, over the years, I guess, um, is this question of horizontal technology versus vertical expertise. Um, and you showed a bunch of, of different, uh, verticals, but, for example, something like medical, Is the claim here that you could apply this fairly quickly, or does it need, do you need a bunch of specialists to come and train the system?
A Um, absolutely, we can apply it quickly to all these different verticals, because that's the beauty of neural networks. It's literally taking a pair of inputs and outputs, and they could be anything. Here we've been showing tags, but they could even be a quality prediction, what's a good quality image, a price prediction for A product on eBay or something like that where, ah, you don't know what to price it at to start. And it can learn these mappings in arbitrary ways. So, the medical, ah, domain might not have categories, the image pixels might be different, but it's just learning that mapping. So, it's just a matter of getting the data, ah, rather than understanding exactly the, the problems that are there.
AI assessment note: “absolutely, we can apply it quickly to all these different verticals”