Dam been writing a book about this and she just TikToked it in 90 seconds.

From ‘The Onion’ Editorial Board:

*stares in livejournal*

My recent perspective on search-assisted AI

I have no doubt that you've worked for Google. It shows in your arrogance. Many, many people who worked at Google do not understand its purpose in society or culture. Google should, actually, always return the correct answer to simple math questions. And you're a fucking idiot for arguing otherwise.

just call it the schumer/jeffries plan to lose the midterms

100% this. Unfortunately, that circle also fits fully inside another circle of people immune to feeling shame about their hypocrisy.

Give me the Woke Bread.

Griffin’s definition of fascism as palingenetic ultranationalism—specifically a reactionary nostalgia for a lost history of the nation that never existed—remains undefeated. A nostalgia for a type of bread that is ubiquitous, created in the 20th century, and rage towards bread as old as civilisation

but for these vile guns

Unrestricted submarine warfare

morning read

Activists and organizers are the real futurists.

Words/phrases Democrats should never use: * bipartisan * my Republican friends * real, valid, legitimate, serious, important (to describe issue raised by Republicans) * yea (when voting on Republican bills) * distraction (to describe bad thing Republicans did) * woke (in derogatory sense)

I never thought I would see this in my life, and that they’re making it immediately generic so it’s affordable is making me want to cry

"We know that a text does not consist of a line of words, releasing a [single meaning], but is a space of many dimensions, in which are wedded and contested various kinds of writing, no one of which is original; the text is a tissue of citations, resulting from the thousand sources of culture."

As someone who has stopped using Twitter ~3 years ago, after having been on it daily for the prior 16 years, it is *so* clear how much the current hateful and extreme state of the platform warps the views of its users, and NOBODY TALKS ABOUT IT. I see the effect so much working in tech.

An important reminder that the state is not the final authority. People, in solidarity, *have real power.*

Sam Altman: I see humans achieving omnipotence using dark matter by… 2030 maybe? That’s if OpenAI is given a quadrillion dollars Theo Von: Yeah Altman: Or, we may become a race of slaves, genetically modified to be harvested by robot spiders. That’s if OpenAI doesn’t get the money Von: Right

Now is the time to think about how to head off the enshittification of the next phase of the internet.

I see grocery/takeout shaming discourse going around again, so this is your periodic historian reminder that cooking from scratch daily while working has never been sustainable, and for millennia everyone either relied on takeout & restaurants or ate in huge households w/ big dorm-like group dining.

And data formats / protocols are literally the only things that matter. Everything in between will become disposable.

If I’m arrested it’s because some peanut on Instagram said @riverriverbooks.bsky.social didn’t read all the manuscripts and I just started swinging first, land SAKES, say it to my FACE

Do y’all want some actual good news? Thanks to a mistake and using an AI image this week on @moistpoetryjournal.bsky.social’s insta (groan, already removed), I discovered that Adobe Express has FINALLY given the filter to exclude Generative AI from the stock photos. See below.

watching you dildo brain cock suckers play w/ AI shit is like watching a dog paw at one of those puzzle boxes that you put treats inside of

When I think about the stupid mistakes LLMs make with words sometimes, (cf. the strawberry problem, fabricating quotations, or mixing up terms etc.), there's a well-known Old English riddle that comes to mind. It's more than 1,000 years old but I think it has something to tell us about AI... 🧵

Being in tech and having a single modicum of critical thinking is just screaming "this isn't what LLMs are designed for" over and over as people shove a bunch of word predictors into critical decision making processes because some glorified used car salesmen told them it would fix all their problems

I watched this (youtu.be/TMoz3gSXBcY?...) on the abuse of LLMs & 'vibe physics' by @acollierastro.bsky.social - well worth a watch! - and I was struck by how the back 15 minutes (the 'thought train') also summed up neatly what it has been like to be a public-facing historian, even before LLMs.1/

"i asked grok" "i asked chatgpt" yeah well i asked carl sagan and he said the greatest enemy of knowledge is not ignorance but the illusion of knowledge 🧪

Every once in awhile, I think of my favorite “…the fuck did you just say?” phraseology, which is from Homer’s Iliad, proving we have never, ever changed as a species, only less fancily inclined. “What nonsense has escaped the barrier of your teeth now?”

I'm having difficulty imagining the scenario where this endeavour, even if real and totally unironically successful, makes back anything close to its investment Is the objective just to capture God first?

We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers. The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.

MechaHitler ruined the term grok the way real Hitler ruined toothbrush mustaches.

This is exactly what the internet is for.

AI slop now making up saccharine Bible verses that do not exist

The reason MechaHitler happened is that LLMs are not sophisticated enough to have a hidden agenda. They can’t secretly “believe” something and consistently pretend like they don’t.

Gurathin said, “I can’t tell if that’s you being passive aggressive or you being willfully obtuse.” I would be more pissed off about him saying that except a) he was right about the passive aggressive thing

The fact that Google, a company ostensibly in the business of "epistemology," continues to run AI results that are wrong in double-digit percentages (to be generous!) continues to boggle my mind. An absolute abdication of any purported value to users.

whenever I hear the government cry about a "budget deficit" it's always just the dril candles tweet. except the candles are cops and tanks.

A lot of takeaways from Zohran’s DSA-powered win but one of them is that liberals are not the problem. The liberals voted Zohran (or Brad and also ranked Zohran). It’s reactionary centrists that ran the disgraceful race-bating campaign.

I’m obsessed with this company that makes shotgun shells full of flower / crop seeds. Very swords to plowshares coded imho.

I'm in the interesting position of being a member of two distinct communities on here, one of which is vociferously anti-AI (mostly my art/philosophy/leftist people) and one of which is generally pro-AI, with varying levels of enthusiasm (my tech community.)

I expect that consumer-facing AI programs will continue to improve and they may become much more useful tools for everyday life in the future. But I think it was a disastrous mistake that today’s models were taught to be convincing before they were taught to be right.

you don’t win a fight with a troll, you only “win” by blocking, muting, and not giving them the attention they desperately want — protect your peace at all costs, it is the most valuable asset

Bluesky is now culture war grounds, & we have to be prepared and adjust for that — aggressively defend your peace and your space here… block, mute, use labelers, limit your interaction settings, do what you gotta do and help each other new clients, new tools, new apps are coming that will help more

LLM-driven software engineering sucks. Also human-driven software engineering (for 95% of teams) sucks. The whole domain is a shitshow. Fad-driven, unempirical, overcomplicated, subject to pathological economics catering mostly to fads and confirmation bias.