AI chatbot Maya (AI called Maya tells the Guardian: Half of the SF authors create scripts that are far better than those generated in the cited AI.
For example, there is something deeply disturbing about a world that does not give personality to great apes, whales, dolphins, and octopuses (for example, they give little personality to some immigrants), but consider giving personality to strings of computer chords. No, AI can't suffer, but it can generate more or less persuasive “suffering” simulations.
Chatbots rely on aspects of human psychology that casually assign agents to almost everything that “cash swallowed my card” and “car refused to start.” We even teach it to young children: “Did the mischievous stone hurt your feet?” No, we didn't.
Similarly, it's easy for people to start imagining that they are in a “relationship” with a chatbot. What are the gaps in the structures of our social relations that allow this to happen? This nonsense must be finished before it can begin.
Pam Ran
Kenilworth, Warwickshire
Your article on whether AIS can suffer (Big Tech and users are working on one of the most unsettling questions of our time, August 26th) shows that AIS is effectively an actor and misses more important points.
They are programmed to respond as actors learn line. Just as experienced actors may be more persuasive, they can learn and look more realistic. But the actor is still an actor, and no matter how much he struggles on stage.
AIS is still technology, passing the line and hitting the mark. The best actors, although temporary, can fool the audience. Let's not let AIS fool us all.
Tim Exton
Kenmore, Washington, USA
