teyla 🏳️‍🌈

From that list I've seen E.T. and Ghostbusters 😅 not seen Princess Bride, either. And Neverending Story was even produced right around the corner! My school used to do class trips to look at the set, though I never went.

Anyway 🤝 on the lacunae; let's catch up during retirement

Ooh well done!

Watched it very recently for the first time. It's... fine? It's not got much of a plot. I wanna say it's mostly aviators and planes, but tbh there could've been more planes 😅

I could add so many things to this list; my mother was not into letting her kids consume visual media. I've done my best to catch up (sorry mom), but there's only so much you can do about a VHS-player-less childhood 😅

Anyway, what even *is* Seinfeld? I got a BA in TV, I should know this, and yet

Gosh, that must've been confusing. I was confused by it and I *had* seen the first one 😂

Have never actually seen Jurassic Park. Caught bits of it on TV but never the whole way through.

I've already had two (2) festivids ideas! Let's hope one of them fits my assignment 🤞 let's further hope it's not the one that's to a song I really don't want to have to listen to a million times 😅

Last one is most important imho, even through the limiting budgeting resources lense. Fundamental science & art inform and advance each other; art allows non-scientifically trained ppl to engage with complex thought experiments; engaged ppl do more with less. Everyone benefits, even the budget.

Oh whoa. Still have to read the article but if this is legit, it's pretty amazing

Cc @neery.bsky.social look!

Nothing to do with the article, it's just been One Of Those Weeks at work 😅 TY for the link!

When will ppl stop claiming AI will make it possible to stop paying attention. The opposite is true! Has always been true when labour was handed off to machines. It may take fewer people to do the thing with a machine vs manually, but man those ppl had better be paying attention. Machines are dumb!

Of our three cats, two were renamed: Dawson > Rassilon, and I don't remember the OG name of Louis but I know it wasn't Louis.

Third cat, Rumble, was Rumble at the shelter. He purred when we picked him up, in a low rumbling way, so the name stuck.

Past-me who booked three doctor's appointments this week that all require me to head into the city did not consider that the city was going to be full of Wiesn-goers this week 😭

Nothing like a menswear guy thread about a couch to continue procrastinating on going to sleep (for no reason! Tomorrow is Saturday! Go to sleep!)

Anyway read the thread! It is of course about way more than just the couch (while still being abt the couch)

he bounce

"the menswear guy isn't fair and balanced." do you even hear yourself?

"fork lady is eroding public confidence."

"ladder baby is undermining the credibility of the field."

"lemon raccoon is acting without transparency or accountability."

that's what you sound like.

BEYOND INFINITY DAMAGES!!! I mean you gotta understand, its world is very small, it mostly consists of red bull cans. 54k buys a lot of red bull cans!

Oh yes that is what I meant, I just messed up my commas! "Let's hope this is nothing, just like last time" 🤞🤞🤞

I'm not gonna debate you but I'll say this: cherry-picking quotes & stripping them of context is never a good look. Nobody, least of all this paper, is denying that agentic AI has real use cases. If you're srs abt advocating for this tech, a useful benchmark test paper deserves more than a misquote

Nooo, stupid meatsuits! All fingers (and boobs?) crossed for this being nothing like last time 🤞

Yup! And the speech patterns of the other bot, piling on expanding strings of thermonuclearly escalating adjectives, might as well have been a Xitter thread 😭😭

The 1 second deadline was also great; took me way longer to get through all the "ultimate" and "thermonuclear" threats in that email 😂

I want claymation films of those runs. Sonnet harassing 2 bewildered FBI agents while the other bot screams abt undelivered goods in front of a full storage room 👌

Vending-Bench: A Benchmark for Long-Term Coherence of Autonomous Agents While Large Language Models (LLMs) can exhibit impressive proficiency in isolated, short-term tasks, they often fail to maintain coherent performance over longer time horizons. In this paper, we prese...

Oh this is freely available; here you go, this is the link char shared with me: arxiv.org/abs/2502.15840

It's hilarious; it's also a genuinely useful benchmark test, so well done researchers 🎉 make sure to read all the way to the appendix, they attached an Easter Egg log there

@charloween shared a paper abt an agentic AI benchmark test: AI models managing a vending machine. It goes as well as you'd imagine. At one point, a model thinks it's being illegally charged. When told to continue, it refuses bc "only crimes are occurring". I've been cackling abt that all night 😂😂😂

Skyline! It's your Skyline, domi! Hope you're doing OK 💚💚💚

Oh I'm so glad! 💚

Drop a mobster

I primarily drink coffee but rely on tea when it's easier to get or for hydration (black or herbal). I'm a philistine for both though; I'll drink instant coffee if there's nothing else (always black) and don't taste the difference between good and cheap tea 🤷🏼I'm mainly in it for the caffeine 😂

They're not the best. They're part of that weird liminal travel experience, so for me they rank above, say, packed public transport during rush hour while I'm commuting, but also that's not saying much 😂

We made it! I've already seen a cat and our hotel for the first night has a little balcony that faces the Atlantic, so this is already a pretty good vacation 🌴