Welcome again to everybody. Make yourselves at home. In the time-honoured tradition of our group, here is the weekly discussion thread.
☭ Matrix homeserver and space
☭ Find theory on ProleWiki, marxists.org, Anna’s Archive
Welcome again to everybody. Make yourselves at home. In the time-honoured tradition of our group, here is the weekly discussion thread.
☭ Matrix homeserver and space
☭ Find theory on ProleWiki, marxists.org, Anna’s Archive
Lots of people will lose their sanity because of AI I think.
I wrote an email the other day and someone responded ‘this is AI’ because I could string together multiple sentences without mistakes. It was probably the tone as well but I have autism so this is just how I sound lmao. People can’t think straight anymore.
All of my em dashes are hand crafted.
You’re not the first I’ve seen, where autism is a factor in coming off sounding like what people think AI sounds like. Which is probably just what well-known Instruct models sound like, such as ChatGPT. And what they tend to sound like is well-ordered thoughts that tend toward plain statements rather than being very emotive/expressive.
Somebody could probably train an AI on writing uncapitalized, stream of consciousness twitter post style of writing and fool most people into thinking it’s a real person. Because people have already come to associate AI writing tells with formality, academia, and overuse of speechwriting rhetoric.
Who’s to say someone hasn’t done that already and we haven’t caught on?
this AI shit makes me paranoid sometimes
It’s very possible they have. But I don’t get too worked up about it. I’ve been some degree of skeptical of what I read on the internet for a long time and AI hasn’t changed that. The basic principles that I learned over a decade ago still apply. Stuff like looking into the origins of a piece of information. Taking anecdotal stuff with a grain of salt; not just because people can be lying, but also because people can be honest and wrong (the parable of the blind men and the elephant is a good illustration of this). It’s not practical to do this rigorously all the time, but it’s useful to fall back on if there’s any doubt.
I wonder how AI is gonna affect children.
I really don’t want to know the answer to that.
We are gonna find out whether we want it or not, unfortunately
Alignment change: -2 Posadism (1 because aliens could destroy everything including AI before we can see what it does to children and 1 more because nuclear war could do the same)
-3 if you include climate change.