“When I talk to experts in the field, the area they’re most concerned about is misinformation. …I think we’re going to have to get better and that means companies coming up with automated tools. It means watermarking videos and other things so that we know they were created artificially. It means having a provenance chain so that you can tell this is footage that was legitimately captured from a device and here’s everything that’s happened to it along the way.” [watch videoOpens in a new window]
— Ina FriedOpens in a new window, chief technology correspondent, AxiosOpens in a new window Opens in a new window
For example, say you have an algorithm trying to predict who america phone number list should get a promotion. And say there was a supermarket chain that, statistically speaking, didn’t promote women as often as men. It might be easier to fix an algorithm than fix the minds of 10,000 store managers.” [watch video]
— Richard SocherOpens in a new window, CEO and founder, You.comOpens in a new window
“A lot of times, the failings are not in AI. They’re human failings, and we’re not willing to address the fact that there isn’t a lot of diversity in the teams building the systems in the first place. And somewhat innocently, they aren’t as thoughtful about balancing training sets to get the thing to work correctly. But then teams let that occur again and again. And you realize, if you’re not thinking about the human problem, then AI isn’t going to solve it for you.” [watch video]
— Vivienne MingOpens in a new window, executive chair and co-founder, Socos LabsOpens in a new window.