BTW, if they'd object and say, "but we do know what these systems do, they predict the most probable next word", then please know that this is false and misleading, because that function is intractable, both to compute & to make, so AI systems cannot be computing such function *at scale*.

Oklahoma Gov. Kevin Stitt said it's too easy for people to feed their families with SNAP benefits and suggested they should have to go to food banks so they feel embarrassed. The real embarrassment is lack of living wages stateline.org/2026/07/09/f... h/t @kelciemmorris.bsky.social @stateline.org

After 100 million plus tokens I have just as much experience with the reality of this technology as any "AI" booster does. Don't fall for the FOMO. The unreliable intern and thought partner modalities of extracting utility from LLMs are dangerous fantasies and prompt engineering is a silly roleplay.

For "AI" to only be an amplifier or an accelerator, it would need to have semantic understanding. You cannot have "this, but more" or "this, but faster" without the system being able to fully process what "this" is.

"As Hadley Wickham puts it, AI is like an amplifier, or a magnifying mirror: it makes you more of what you already are." This is a line I've seen occasionally from "AI literacy" sources and sometimes even seemingly critical sources: "AI" is an amplifier, or an accelerator, or some similar framing.

Guess who is the only guy in town who now has a copy of a judicial warrant proving that the sheriffs department lied for DHS 🤠

Amendment 5 will financially bankrupt Missouri to compliment MAGA's morally bankrupt race to the bottom. Most rural newspapers will carry either only complimentary comments from MAGA or no info on proposal. St. Louis Post-Dispatch article via Wayback Machine archive.today/2026.07.10-1...

FUCK FUCK FUCK FUCK NOOOOO https://ca.pcmag.com/news/16760/fcc-approves-reflect-orbitals-giant-mirror-satellite-that-astronomers-hate

What if I told you one woman, with little accountability, is wearing the name of the United Nations to launder false anti-trans forum posts into global legal fights, including the US Supreme Court? Our reporter S. Baum investigates Reem Alsalem in our collaboration piece with The Nation.

U.S. troops aren’t supposed to be policing an American city just because the president thinks it’s enemy territory. There’s no reason for this guard deployment to Memphis and definitely no reason for them to be shooting or killing ANYONE there Out

I think I’d let scissor lady step on my throat

"When Stephen Miller or other MAGA types suggest banning pregnant tourists from visiting the US, this is a view of children not only as objects but as essentially a material resource in a race war." @aselrod.bsky.social is good in digging out the deep and sinister narratives underneath MAGA tropes:

If you’re not inclined to read a Supreme Court decision, I recommend this article scholars.law.unlv.edu/cgi/viewcont...

That’s right, it’s Lassiter v. Department of Social Services, where the Supreme Court decided—against all logic, reason, and precedent—that it’s fine to terminate a parent’s rights to their child, permanently severing the relationship, without the parent having a lawyer www.oyez.org/cases/1980/7...

All of those are constitutional rights in criminal court. None of them are constitutional rights in civil court. And family court is civil. They can take your children away and terminate your parental rights, but you have fewer constitutional protections than someone charged with burglary.

I'm been wondering if it would come to this and apparently it has. huge move from cloudflare, which hosts a fifth of all websites on earth.

Please consider not saying that some AI models or some statistical models are good, useful, beneficial, whatever else positive, in science if it's not your research area. It's ok not to know things. It's ok not to both side things. It's ok to ask colleagues who are experts.

If you're anti AI & making claims that some of the models under the AI umbrella are ok because they seem to be serving scientific goals, then you fell for a type of hype. One from the 2010s; stats models aren't blessed by their use in science and they can erode our thinking. bsky.app/profile/oliv...

how to keep wildlife going during the heatwave and drought we are having in the UK www.littlegreenspace.org.uk/features/hel...

Gift link for this excellent article from the indomitable Chris Gottlieb about, as she puts it, “the largest, most invasive arm of government that systematically flouts constitutional protections without challenge” slate.com/news-and-pol...

Quick tutorial on how to opt out of Meta's new feature which allows anyone to tag your Instagram handle to generate AI images of you. @evystadium.bsky.social on Muse Image in 40 seconds:

If these rules are accepted and take effect, political appointees will be able to decide what science does and doesn't receive funding, with scientific peer review being merely "advisory." sciencebasedmedicine.org/lysenkoism-2...

Read the new research in Antiquity 🆓 Haughey’s Fort: a major complex of power, production and ritual in Late Bronze Age Europe - James O’Driscoll & Patrick Gleeson doi.org/10.15184/aqy... 12/12 🧵

Marx and Engels published the Communist Manifesto 168 years ago. It was specifically a reaction to the Industrial Revolution.

if you want an idea for how insane midwest politics are becoming but also how DSA (brand) is clearly seen as an asset even by right wingers, then look to Joplin, Missouri, where a lead for Americans For Prosperity attended a DSA meeting to lie about their positions on the GOP astroturfed amendments

you don't even have to say "better things are possible" to be mocked and tut-tutted anymore. you just have to say "things that literally existed and worked fine a couple years ago are still possible."

I feel people do not get why countries make these laws. > Peru lawmakers push open-source software Peru was seeing it from the mid 2000s. Yes, they banned Microsoft.

I’m in State government IT and the direction we’re heading is alarming. Not just AI pushes but shifting everything over to the cloud and various private vendors. We have smart people and we even own our data center facilities outright. The state can and should maintain autonomy.

I've pulled this thread into a post, and I've also covered the current state of the CHARLIE Act. bookriot.com/federal-book...

People love age-verification because they think it just affects kids. But kids don't have IDs. And there are prohibitions on collecting their data. It's the adults, not the kids, who have to submit a face scan, upload a government ID or provide other personally identifying information.

Get Loud: Public Library and School Internet Access Is Under Threat The FCC wants to change the e-Rate program, which could eliminate discounts libraries and schools get to provide internet to their community. bookriot.com/proposed-fcc...

So let's not pretend learning from the AI bubble is for the future, as it plays into the hands of the technology industry and the AI ideology specifically. bsky.app/profile/oliv... 7/

Reading is a great place to start. The AI bubble is based on funding from billionaires & us believing their hype. This second we can dismantle within our communities as a first step towards also dismantling the first. I have some resources here: olivia.science/ai 8/ bsky.app/profile/oliv...

It's fascinating to live in a time where the majority of global consumers are just fed up and hate everything new on the market. Hate removal of physical media Hate forced AI Hate video surveillance in cars and homes Hate micro transactions Hate digital subscriptions What a time to be alive!

Anyway I think the moral rot of our society is probably not unrelated to the disdain for the workers who effectively raise our children, take care of our elders, clean our streets, etc

Is this good

This story has been percolating in local media since last month – credit to Syracuse.com, which appears to be the lead dog on this story — but it just hit the national radar over the last 24 hours or so. talkingpointsmemo.com/morning-memo...

The best edit I’ve seen in weeks, you absolutely have to turn the sound on for this one.

I have edited a lot of sexual-assault-accusation articles, and this one easily clears every veracity check: contemporaneous emails with her therapist, a boyfriend who knew, a woman she warned—all before Platner ran for office. Good on Politico for nailing it, and on Jenny Racicot for speaking up.

NEW: The ICE office that customarily probes employee misconduct is being refocused to investigate ICE critics, opening hundreds of "doxing and threat" cases and subpoenaing tech companies to unmask anonymous posters. Detention facility inspections, meanwhile, are plummeting. by @regret.bsky.social

A KCUR investigation discovered Lenexa resident Canyen Ashworth was tracked by the Lenexa Police's license plate readers after he wrote a column critical of the city’s police. “My experience with the Lenexa PD has, I would say, radicalized me against these automated license plate readers."

New in PN with @juddlegum.bsky.social: How Kalshi infects the news "Since last December, both CNBC & CNN have promoted Kalshi to viewers extensively, frequently vouching for its accuracy. The existence of financial ties between the networks & Kalshi, however, is disclosed to viewers inconsistently"

While scientists have long braced for climate change, the growing severity of its impacts is shocking them. www.bloomberg.com/graphics/202...

"In nearly every state, the official humanities nonprofits created by a congress to help make history and literature accessible to all Americans had to give up their 250th anniversary planning when DOGE pulled their federal funding..." www.notus.org/agencies/dog...

Similar to genAI writing. No matter the skill of the user, AI returns the opposite of good writing - jargon-filled, lengthy sentences of corporate-speak with no originality, no audience focus nor resonance, and no cited sources. Too much time and trouble to clean up that mess.

“Home security cameras are often marketed as a form of community connection, but they are ultimately used as a means of isolation and community policing — and the negative consequences disproportionately impact minorities…”

Some people have been saying.

“AI can give you a polished version of your own half-formed thought,” said Duncan Brumby, a professor of human-computer interaction at University College London. “The danger is that the polish comes by sanding off the distinctive edges of what you actually meant.”

Paper is here:

Academics “examined the behaviour of mainstream large language models provided by Elon Musk’s xAI, Meta, Google, China’s Alibaba and France’s Mistral and found the introduction of bias happens even when the AI tool is instructed to preserve the original sense.”