Why regulating AI is essential but might be impossible

[MUSIC] AI is becoming more and more pervasive in our lives. Again. It's in our back pocket, it's in our homes on the kitchen counter and it's in the cars that will be driving us to work or to the store, it's going to be a present in more and more and more aspects of our lives. And so it's really important that lawmakers, policymakers, that they understand the technology. It's moving extremely quickly. [MUSIC] Whatever is created really needs to be mindful of the context, and as well as what the particular risks are. So again, they can be mitigated in a common sense way. Having one overarching, Policy for all of AI treats AI as if it's this monolithic technology, and it's not. It's exceptionally diverse and complex. So it needs to be regulated in that same kind of manner. [mmm] I think it's tricky to talk about a regulation in the abstract because they I there's always a plus x. It's always applying to some particular industry. And so I do think it makes sense to regulate AI as it applies. To ginormous, as it applies to transportation. Transportation authorities not letting any start up just drive their self-driving car along the road without a lot of really hard safety tests. Doesn't make sense to just allow a eye to make important life or deft decisions in medicine without reading carefully about making sure it's not just biased and working for a particular subset of people. And it also makes sense to apply regulations when it comes to Military. It's actually something I don't work on and I hope it can be very much regulated. Cuz we don't wanna have autonomous weapon systems that make death decisions by themselves. So I think AI regulation makes a lot of sense when you look at each particular industry. It doesn't make sense to be in the abstract. [MUSIC].