(googling 'Island of Inchkeith') huh, three one-star ratings but nobody wrote a review

Looking at the language deprivation wikipedia article & popping open the "Reception" accordion to see if people liked it or not

Did this at a party once with a mango. Also they had specifically taught me how to cut a mango at the last party.

Every day, she just sits and waits for the balloon. Toward the end of this journey, Mr Beast, in narration, says something like, “she does this because if she doesn’t hit the balloon, her living space will be smaller. But also because it’s the only thing she has to do.”

Technically the difference is that if I like it, & it's good, then it's ML. But if it's bad? Mmmm probably AI.

whoa, what a pretty figure

personally I can’t believe there’s so many leftists who, when presented the “automated” part of fully automated luxury gay space communism decided it was actually immoral to automate things

I think this is a bad way to think about social media

This doesn’t seem right

People talking about "LLMs can't make new information" is so fucking funny. Ah yeah, AlphaZero hasn't found new chess strategies. Protein folding models don't make information.

So based

🚨 The DOJ appears to have redacted Donald Trump’s name from the allegations made in this exhibit in the Epstein files. Trump’s name was in the original release. Now, it’s blacked out. See for yourself.

Which Ann Beattie collection should I read?

I want topic-based social media so badly!!!! but not badly enough to read reddit

I'm on my knees

I think a key problem is that to most critics, it must be everything. It has to destroy the environment, eliminate every job, enable a fascist takeover, melt our brains, all while being a uninspired, unremarkable piece of boring technology that also doesn't work.

Really interesting research showing that while ultra-specialization in a single discipline might lead to better results early in one’s career, multi-discipline training and practice pays off big time in the long run. This applies to a range of professions from scientists to athletes and more

The Virgin Spanking the Christ Child before Three Witnesses: Andre Breton, Paul Eluard, and the Painter Max Ernst, by Max Ernst, 1926

Larian produces very good work. If they say they can integrate AI into their workflow as part of doing that, all evidence I have right now suggests that it works for them. That doesn't mean it works for your situation or mine, but I assume it works for them until I get evidence otherwise.

i was trying to figure out what the thing the tower in Pathologic 2 reminded me of. claude is such a weird fucking product

I need you to know that this is the most normal facebook marketplace listing at the top of my feed.

Decided c2pa is probably silly shit from first principles but now I have to actually read about it so I can say it's bullshit from second principles

this is going to be an ongoing problem in the post-Trump world

Removing the context because I think "this technology will go away" is detached from reality, but I think there's def some truth to this, which suggests some of the more interesting avenues for generative models (e.g. interaction).

HUGE NEWS from the new Epstein files: Records show that Maria Farmer, who worked for Epstein, filed a "child pornography" report to the FBI in 1996. The FBI has never before acknowledged that complaint. The case went nowhere, and mass abuse followed. www.nytimes.com/2025/12/19/u...

Oops, all redacted

Google released the open-weights of T5Gemma 2, an encoder-decoder models Built on top of Gemma 3, T5Gemma 2 is a multimodal, long-context, and heavily multilingual (140 languages) encoder-decoder model, where most models today are decoder-only.

So that means AI is basically responsible for about 0.05-0.13% of global emissions. That’s right. At most it’s around a tenth of a percent.

I mean this literally, your average person thinking ChatGPT is a weird sex thing would be an improvement on the average person thinking they know how to use it. I am a smart, educated person and I know just enough to know *I* do not know how to use it nor would I benefit from trying to learn.

Lmao awesome to see people posting about how cool it is that we can "live like europe with gdpr" if we kill section 230 on the site that will immediately blow up if we do that

Things are going great over at X, The Everything App

fall into a pit and die

a lot of random semantic disputes about "generative" ai seem to be people wanting there to be some way in which the companies and products that they don't like are doing something categorically different from those that aren't it's just really not categorically different

Stupid fucking country

I am learning now for the first time that the little weatherman who lives inside my tv set was *not* doing his own forecasts and was in fact cheating by using computers to do it for him

I'm finally at the acceptance stage of seeing people say ML and LLMs are too very very different separate things lol

The Dispossessed

Stay safe out there y'all

MY 4-YEAR OLD: *calmly, while coloring a picture of the Grinch* If baby grinches die, we can eat them. If they’re clean. That’s ok

idk, I think Firefox giving folks the ability to run ai models locally on their devices is fine, and significantly better than corralling their users into some bullshit cloud service.

you may not like it but this is what optimal humanoid motion looks like (vid from junzhe he)

Someone's gotta generate all deadly sins as disney princesses or the domer meme, realest stress tests

It's weird how many people think they have beef with a whole field when they have beef with specifically like 5 companies

there are certainly things I don't like about AI/LLMs but a tremendous amount of 'criticism' is just dumb clickbait slop

This is also very likely a result of annotation workers for things like RLHF often being from Kenya in particular.

Yellow-Red-Blue, by Wassily Kandinsky, 1925 (detail, rotated), 📸 by @riogiancarlo.bsky.social