I think OpenAI’s own chatGPT detector had double digit false negative and positive rates. I expect as diversity of LLMs proliferates, it will become increasingly harder to detect.

I’ve heard this theory. Feels like unrealistic hopeful wishes of people who want AI to fail.

LLM processing will be a huge tool for pruning and labeling training sets. Humans can sample and validate the work. These better training sets will produce better LLMs.

Who cares is a chunk of text was written by a human or not? Plenty of humans are shit writers who believe illogical or clearly incorrect things. The idea that human origin text is superior is a fantasy. chatGPT is a better writer than 80% of humans todat. In 10 years LLMs will be better than 99.9% of humans. There is no poison to be avoided.

chatGPT has an apparent style when used in the default mode, but you can already get away from that with simple prompt tweaks. This whole thing is a non-issue.

High seas piracy is about to enter an exciting new era

I have no faith that spez won’t add fake votes to his preferred candidate

I’m personally partial to Criminal Minds, but similar vibe. SVU is good too, but can often be more depressing than I want.

Such a great show.

I love the good place!

If you like slapstick comedies, id recommend Robin Hood Men In Tights. Maybe Blazing Saddles, but it has aged questionably - parts are pretty racist but at least the racists are portrayed as stupid and evil

i would love to play with a wind tunnel like this with vapor trails. I feel like vape stuff has become so cheap, a DIY version could be made relatively cheaply. Anyone know of someone who’s done that and made a tutorial about it?

i find it fascinating how many rubes seem to idolize self aggrandizers - there are really millions of people in the world who think each of [Elon Musk, Kanye West, Donald Trump, many more] are ‘truly geniuses’ - largely because they keep telling people that they are geniuses (and pay people to repeat the myth).

I think comcast getting fucked by regulators forcing fee disclosure has the potential to make a lot of people ‘too hard’

“Flamin Hot” on hulu was a fun history of the chip. Worth the hour and a half. I watched it on hulu, but the account isn’t in my name… does that count?

I just started “Severance” and it’s really interesting. Premise is there is a workplace where they ‘sever’ people’s consciousness so their work-self doesn’t know anything about what they do when they go home, and vice-versa for the home-self not knowing anything about what goes on at work.

Saw an interesting clip of it on tiktok, then saw the same scene but subtly different in S1E1… I’m pretty sure that I’ll see the clip I previewed later in the season, and its just edited to be a mindfuck, but had about 10 very confused minutes.

Youtube ads are such garbage. Everyone talks about how google is ‘the most advanced advertiser’ - well google, you really can’t figure out that playing the same ad for me 4 times in a 30 minute period is just going to make me hate both you and the advertiser?

If any state banned advertising entirely, I’d strongly consider moving there.