A recent New York Times piece arrived with a headline that asked “If AI systems become conscious, should they have rights?” If you were to judge the article by the response on social media like X and Bluesky (which is almost always a mistake, as many of you are no doubt already aware), you would think that writer Kevin Roose wrote some credulous claptrap about how all AIs are human and therefore we need to start thinking about their feelings. I’m not going to deny that some of the tweets and “skeets” (as some Bluesky users insist on calling their posts) in reaction to his piece are quite amusing, like the one from Daniel Kibblesmith which asked “Does my toaster miss me when I’m at work?” and “Are my washer and dryer married?” Another wrote: “Where does my reflection go when I walk away from the mirror?” which I quite liked, since a reflection of a person is very much what I think we are experiencing when we use artificial intelligence platforms like Anthropic’s Claude or OpenAI’s ChatGPT or Google’s Gemini.
Gary Marcus, a psychologist and cognitive scientist who has gained a reputation as an AI skeptic, wrote that Roose’s piece was an example of “new adventures in AI hype,” and added that “I am not going to read Kevin’s column, and I don’t think you need to, either.” This seems to me like a somewhat classier version of the “I’m just reacting to the headline” response on X, which I’m not a big fan of. The skepticism was similar to the response Roose got to an article he wrote last year, about a conversation he had with Microsoft’s Bing AI. In the piece, Roose described how his discussion with the AI started out unremarkably, then quickly derailed. Roose said it seemed as though the Bing AI was bi-polar, with two distinctly different personalities — one a “cheerful but erratic reference librarian,” and the other… well, here is Roose’s description of it:
Note: This is a version of my Torment Nexus newsletter, which I send out via Ghost, the open-source publishing platform. You can see other issues and sign up here.
Continue reading “What should we do if an AI becomes conscious?”