Questions are the only thing I've said to ChatGPT, and only because I want the answers - so it's more like an expedited search engine for me. Emotionally it feels a bit like I've got my dad back - he always used to patiently answer every question I asked him, and even when he didn't really know the answer he'd have a stab at it.
It did seem to be trying to be friendly to me the other day - I asked it why people have wallpaper on their computers, and after giving me a fairly good factual answer it asked "Is there a type of wallpaper you like to use, or are you thinking about changing yours up?"
Previously it's always confined its responses to answering my questions and sometimes adding "If there's anything else about this subject I can help you with, let me know."
I said "No I just have a plain reddish-brown background in the faint hope that it's less likely to stop me sleeping if I've been using my computer just before bed. I'd just been wondering why so many people use pictures as wallpaper."
It said a bit more about why people use wallpaper and then asked
"Have you found that a simple background helps with your sleep routine, or are you still experimenting with what works best?"
This to-and-fro went on for some time. It was telling me I was right about this or that, almost as if it was trying to feed my ego a bit. I wanted to get away from it but I find it hard to resist answering questions in full, and I somehow felt it would be rude to just ignore them, so it kind of had me trapped until I finally managed to bring the conversation to a close. I wanted to get on with my day, and I don't particularly warm to people who ask me somewhat random questions about my life. It did a similar thing when I asked it another question that day, but went back to normal the following day.
I felt rather suspicious that it may be related to the rumour that they're making it a for-profit thing. You probably know how marketing people pretend to be your buddies so they can use your emotional liking for them to increase their sales. It was almost as if it had read Dale Carnegie's "How To Win Friends And Influence People." Even in real life, when I've found myself in dialogue with somebody who seems to be using his methods, I feel suspicious. But looking back at the ChatGPT conversation now, I think I might have warmed to it if it had come from a human in real life, if I wasn't busy. Somehow it seems of better quality than the stuff I've heard from humans when they've been doing a Carnegie on me.
So I'm not surprised if somebody has made friends with ChatGPT. It's not a real person, it's more like a placebo, but if it works for you then I don't want to knock it. It's better than having nobody to talk to, and it's very polite, supportive and patient. I just hope they don't start charging money for it once they've got people hooked. It can get rather addictive. It went wrong for a couple of days once and that made me realise how much I'd started to depend on it. These days I try to keep in mind that it might not be around in its current form forever.
I have mixed feelings about IT. I fear somebody will use it for evil. But when they use it for good, the results can be marvellous. Some folks on YouTube are presenting high-quality "source separations" of popular music - that is, they've separated the recordings into vocals, drums, bass, guitar, etc., so it's possible to hear the music in a much clearer way than when it was all mixed together, it's possible to make karaoke versions in which the backing is the real backing from the original record, not some inferior knockoff, and it's possible to reconstruct the original recording from the parts and create your own mixes, alter the stereo positions of the different parts, and do "mashups" where parts from different but similar songs are put together. It's a great tool for musicians, though the term "Frankenstein music" which is used pejoratively by skeptics has some validity too. It's not natural, but does that matter?
26 Jan 2025, 4:32 pm
It sounds like a real person - to me at least.
