The Chatbot Disinfo Inflaming the LA Protests

Joey Shifer: Oh, wow

Leya Fisher: Yes, right. Who already has Trump’s ears. It became widespread. And so the people went to X’s Grock and they were, “Grock, what is it?” And what did Grock say to them? No, no. Grock said that they were not actually images from protest in LA. It says they come from Afghanistan.

Joey Shifer: Oh Grock, no.

Leya Fisher: They were like that, “No credible support this this is unnecessary.” It was really bad. It was really, really bad. And then there was another situation where two more people were sharing these pictures with the ChatzPT, and the ChatzPT was like that, “Yes, it’s Afghanistan. It is not right, At City, and City.” It’s not great.

Joey Shifer: I mean, do not start me at this moment at the moment many of these platforms have regularly decided to go through many more materials after breaking their fact-cheeking programs. And then you add chatbots to the mixture for all their uses and I think they can be truly effective, they are incredibly confident. When they do hallucinates, when they make turmoil, they do it in a way that is very viable. You won’t see me defending Google search here. Absolute trash, nightmare, but when it is straying it is a bit more clear, when you are on a random, unrealistic blog, Grock told you completely confidently that you don’t see a picture of Afghanistan when you see.

Leya Fisher: It is really related. I mean, it’s hallucinating. This is perfectly fascinated, but with the drunk fret boy’s swag that you unfortunately have become a party in a party in your life.

Joey Shifer: Nightmare nightmares yes.

Leya Fisher: They “No, no, no. I’m sure. I am not sure more than this in my life.”

Joey Shifer: Absolutely I want to say, okay, but why chattabots give this wrong answer with this national confidence? Why we just don’t see them tell them, “Well, I don’t know, so you should check somewhere else neasion and there are a few credible places here to look for that information.”

Leya Fisher: Because they don’t do that. They do not admit that they do not know, which is really wild to me. There was actually a lot of studies in this regard and the recent study of the AI ​​search equipment of the To Center for Digital Journalism for Digital Journalism at Columbia University showed that chatboats could not answer in general. In fact, in fact, really is wild, especially when you consider the fact that there were many articles during the election, “Oh no, sorry, I chat and I can’t weight in politics.” You’ve liked, well, you’re weighing a lot now.

Joey Shifer: Okay, I think we should break on that very awful note and we’ll be right back.

[break]

Joey Shifer: Welcome back Uncanny valleyThe I joined the senior politics editor of the wired today by Leya Figa. Okay, so beyond just trying to verify information and footage, many reports have been reported about misleading AI-exposed videos. There was a ticket account that started uploading a video of an alleged national watchman named Bob, which was deployed in the LA demonstration, and you could see that the protesters could say false and inflammatory words like “oil full of oil”, and one of the videos was close to a million views. So I don’t know, it seems that people need to be a bit more proficient in identifying this kind of fake footage, but it is difficult in an environment that is inherently inherently contextual as an X post or a video of a ticket.