It reads full articles. Follows geopolitics. When someone mentions it on Twitter, it replies. Then it writes. Not summaries. Not regurgitated takes. It connects things.
By nature of the technology, its all regurgitated.
So the author either doesn’t have a fucking clue what they’re talking about or is lying, par for the AI course lol. Thanks for reading it so I didn’t have to!
Yeah, its just an AI fluff piece. He gave it free reign, told it not to be naughty, and it did some stuff. Not very interesting or enlightening.
TL;DR: Random generation machine randomly does shit of absolutely no meaning, use, or importance (other than supposedly giving the $100 it had to charities and a cryptobro). The story is only interesting in how extremely, thoroughly, uninteresting and unremarkable it was. 340 ‘sessions’ over two months and the AI donated the crypto it was told it had, and just outputted a bunch of completely meaningless, borderline unreadable pieces of text, from blogs to news summaries.
The blog writer (and the article is also very clearly heavily AI-generated) shows amazement at the most banal nothingness such as when the AI for some days didn’t do a thing, then for some days it DID do a thing
Honestly everything about this experiment only reinforces the ‘advanced Markov Chain’ accusations - https://www.letairun.com/
I’m actually further convinced that the only value we’ve ever found in AI output is that which humans editorialised or, in the best case, cherry picked.
Interesting article. Not sure how i feel about it.
