reporter: and that s the greatest fear of all. spreading misinformation. chatgpt, a tool designed to help humanity, could ultimately hurt it. people who want to manipulate elections and things like that, instead of like writing one thing at a time, you could write thousands of things to give, for example, vaccine denialism more oxygen than it deserves. can with we just go back to the 80s? it freaks me out. you know, my thing is, look, i am all about the future being now we have to roll with the changes and stuff, but how reliant are we going to get on a.i.? like they said, it s not perfect. now you have a.i. checking other a.i. and perhaps even spreading more misinformation. i don t know about that. it s terrifying. this is going to disrupt so many industries. it is going to remake the
in a sense, yeah. but can it spot misinformation? oh, okay, yeah, so as opposed to misinformation, it is more of like it could only spot if something is ai generated or human generated. and that is the biggest fear of all. spreading misinformation. designed to help humanity, could ultimately hurt it. people that want to manipulate elections, instead of writing one things at a time, you could write thousands of things to give vaccine denial optionism more oxygen than it deserved. and with me now is sarah swisher, and host of the podcast on with kara swisher. cara, let s start with the potential for chat gpt to be used in cheating, for example we tried a 2000 word essay on climate change and they spit out a full essay and some schools are starting to ban it.
it s more of, like, can only spot if something is ai generated or human generated. reporter: that s the greatest fear of all. spreading misinformation. chatgpt, a tool designed to help humanity, could ultimately hurt it. people who want to manipulate elections and things like that. instead of, like, writing one thing at a time, able to write thousands of things to give, for example, vaccine denialism more oxygen than it deserves. reporter: and gary marcus, you just heard from there has written books about artificial intelligence. founded companies about artificial intelligence says we re about 75 years away from ai truly being human-like. until then we need regulation and representative ted lu from california plans to introduce legislation to create an ai commission to oversee all of this new technology. good idea. we know a lot but also very little about it. people are passing college courses with papers written by
cross-checking another ai? in a sense, yeah. but can it spot misinformation? oh, okay. yeah. so as opposed to misinformation, it is more of like it could only spot if something is ai generated or human generated. and that is the greatest fear of all. spreading misinformation. chat gpt, a tool designed to help humanity could ultimately hurt it. people who want to manipulate elections and things like that, instead of writing one thing at a time. you could write thousands of things to give for example vaccine denialism more oxygen than this deserves. and gary marcus went on to say that we re about 75 years away from the place where artificial intelligence like chat gpt is human like. so until then, he said that we need regulation and just this week representative ted lieu