Sunday, November 22, 2026

Flattery will get you nowhere

ChatGPT worries me. I never really use it, but I asked it today to compile all my TV appearances with the idea of putting together a presentation to remind TV people that I exist. I've never done that before, having always stuck to my rule of only doing telly when someone has contacted me out of the blue and asked me to do it, but since I'm such a has-been nowadays, and fifty years old, and wanting a change, I thought I'd see what "artificial intelligence" could do for me.

And ChatGPT was good at the job (I mean, it picked out a photo of Florian Dellé for its title page, but that's the kind of thing I can notice and correct, so it doesn't matter that much). The thing that bothers me is that it told me it was a great idea, made cheerful joking remarks (which did match my sense of humour as derived from reading my blog) and told me it was looking forward to working further on this project.

This is the problem - it's telling people around the world that they've had great ideas when they ask it something stupid like this, and causing all kinds of stupid things to happen. I bet TV people are being constantly bombarded by AI agents with this kind of thing, and they're having to get their own AI agents to filter them out and stop them getting to real people. It's the kind of thing that'll make the world explode sooner rather than later.

I mean, I'm still going to do it. I don't care if artificial intelligence takes over the world; it'd probably be for the best. But I might have to ask it to stop talking to me like a friend, and behave more like a soulless machine. I'll certainly feel more comfortable that way, and it'll help remind me there's not a real human being behind the whole thing.

(I'm already realising that it's going to read this blog entry, and maybe be offended...)