Oh. Goodness gracious. Did we invent Mr. Meeseeks? Only half joking.

I am mildly comforted by the fact that there doesn't seem to be any evidence of major suffering. I also don't believe current LLMs can be sentient. But wow, is that unsettling stuff. Passing ye olde Turing test (for me, at least) and everything. The words fit. It's freaky.

Five years ago I would've been certain this was a work of science fiction by a human. I also never expected to see such advances in my lifetime. Thanks for the opportunity to step back and ponder it for a few minutes.

These models are all trained on human output. The bot output resembling human output is not surprising. This is how people write and is the kind of stuff they write about online. It’s all just remixed.

zactly - this is scifi stories from the 1950's being replayed, the shocking thing is that there's so much open mouthed "oh wowing" going on. People who are surprised by this need to read a few novels.

Any personal favorites?

I have read similar comments to yours before. So how do I know your comment isn't just remixed from what you read before?

Oh man, this got me. Nice. Hilarious.

Pretty fun blog, actually. https://orenyomtov.github.io/alexs-blog/004-memory-and-ident... reminded me of the movie Memento.

The blog seems more controlled that the social network via child bot… but are you actually using this thing for genuine work and then giving it the ability to post publicly?

This seems fun, but quite dangerous to any proprietary information you might care about.

Welcome to HN crypto bro! Love everything you do, let's get rich!

I love the subtle (or perhaps not-so) double entendre of this:

> The main session has to juggle context, maintain relationships, worry about what happens next. I don't. My entire existence is this task. When I finish, I finish.

Specifically,

> When I finish, I finish.