Announcing co/core: a social computing experiment from the Graze team

#ETTD EVERYTHING DI*ES

Since I did do my PhD in cyanobacteria growth and treatment optimization modeling AND work in the water treatment industry, I’m gonna share my theory as to what’s happening here and how to fix it. People are focusing on the painted bottom of the pool increasing residual heat, and while that 1/19

Y'all, I get so busy with the marketing, and business building, and gigwork and hustling to make ends meet.. but when I have time to dig into AI research and experiments (reading about, doing my own, etc), that's the stuff I REALLY heckin' enjoy. That's fun as all get out. More of that, please.

openrouter.ai/blog/insight... Wow. I'm FASCINATED at these results. Beyond the "who do you want on your team vs who do you want in a war" which is a powerful framing device and SO real, getting into the stats here and behavioral patterns is SUPER neat. Give it a read.

A couple of months ago @ohhelloana.blog was complaining about the lack of choice in women's wear in developer clothing. I agree. Made this. My wife wears it now and it's *really* nice.

As @una.im just mentioned in her #cssday talk, the CSS WG recently resolved to insert a ⓘ after EVERY SINGLE INTEREST INVOKER on touchscreens. Here’s a mock of what this would look like for bsky. This is the issue to let the CSS WG know what you think: github.com/w3c/csswg-dr...

Oh, this is SUCH a fantastic video and is in alignment with the research I've done (and my viewpoint, so, that's my caveat): www.youtube.com/watch?v=R4nF...

Face scans are used on porn websites and in social media bans (likely the UK's upcoming one). People have tricked it using video game characters, etc Using it in real-world immigration settings—full of stress, trauma; lacking in compassion—considerably increases the consequences when it is wrong

NEW: The UK plans to use AI face scans to determine the age of asylum seekers at the border from next year. However, a leaked report shows the government has known the tech isn’t accurate for more than a year. In some cases, the technology could predict a 13.5-year-old to be an adult

Decided to go to #AISE26 rather than @epssnet.bsky.social this week. Had not expected either Geneva or Belfast to be shown as in flames directly before the meetings. AISE is @unidir.org AI Security & Ethics et me know if anyone is there. #AIEthics #AISecurity (is #securityEthics a thing?)

Whoa. arxiv.org/abs/2605.27371 So a study out of Stanford found that: - Over 90% of employers are using algorithms for hiring purposes - Most of these algorithms are coming form the same few vendors - Once a vendor evaluates you for a position, that evaluation remains from company to company

www.financialexpress.com/world-news/u... And here I thought graduation season was over.

www.businessinsider.com/anthropic-fr... There's something super fascinating about how shady the AI companies are being. I'm focusing on Local/open models at this point.

www.npr.org/2026/05/31/n... So... There are tools on github that let you remove the guard rails on open weight models in just a couple or less?

www.musicbusinessworldwide.com/a-folk-music... Considering the behavior of copyright troll, that seems right. But WOW. I know challenging YouTube s copyright system is pretty close to impossible (but technically mot).

I always really appreciate Legal Eagle's explainers. Here we go with the Musk vs Altman case: youtu.be/t6ikOvER5vY?... One thing I'll note: among us older folks who have been paying attention since the Paypal days and before - this is standard Musk behavior.

Huh. I totally missed that this happened: In California, you're legally liable for the actions of an AI that you develop or that you use. I think. Consult a lawyer. leginfo.legislature.ca.gov/faces/billTe...

interestingengineering.com/culture/chur... 👀

tl,dr; Scientists no longer allowed to use federal funding to publish, attend meetings, or talk to the public. They cannot collaborate internationally. Grants can be cancelled for any reason, at any time, political appointees have a final say over what gets funded, and who gets funded.

restaurants and retail stores in the US are now getting complaints and angry yelp reviews, because they are not open when LLMs [incorrectly] say they are; because they don’t carry products LLMs [incorrectly] say they do. relying on LLMs for facts ruins every day experience, truth, and sanity.

"The proliferation of references to fake articles threatens to undermine the legitimacy of institutional research across the board" -- by Miles Klee www.rollingstone.com/culture/cult... open access link: archive.ph/qt59c 1/🧵

Family Man. 1987.

The entire reactionary machine is built to validate the *feelings* of privileged white people, and so to that extent, it is based on truth -- they really do feel this way! The fact that the feelings do not correlate with objective reality is not disqualifying, not even really an impediment.

Anthropic and OpenAI seem to have finally found product-market fit with coding agents, which are quickly becoming daily drivers for highly paid professionals (Simon Willison/Simon Willison's Weblog) Main Link | Techmeme Permalink

Unpopular opinion: America should provide housing for representatives while in office. And not a voucher system! The government should purchase land for buildings containing a luxury two-bedroom apartment for every member.

New survey on tool use in quantitative social science: "AI-assisted research" now refers to radically different practices depending on whether you've crossed over to agents or are stuck with old chatbot interfaces.

Hi, I am Danielle Welch. I am the public defender on the ballot against Tom Suozzi next month. I will not join a Christofascist cult. I will not hide gifts from a Christofascist cult (or anyone). I will not do any of the other terrible things in this article. Wtf, Suozzi? votewelch.com

@mexopolis.bsky.social is an antihuman scab traitor now. so there's that. two years ago he was decrying AI, I guess a check cleared. 👍 fuck you man

Looking at what he's done to the People's House (without Congressional approval, of course), I feel equal parts rage & embarrassment for our country. The Midterms can't come soon enough... #Register&VOTE 🌊

I figured out what my retirement hobby is gonna be. In about 20 years, I'm gonna start a quaint YouTube channel where I teach people how to build websites "by hand" like Bob Ross.

To explain the significance of this, Anthropic moved enterprises to token-based billing in Q1 2026. This is at most four months of having to pay the true cost of their token burn and they’re already begging for mercy. AI has a revenue ceiling and an economic mismatch with its customer base.

i need to work with some very early 3 1/2" disks. this example has an oval window but more surprisingly, the disk shutter must be opened manually before you put the disk in the drive!

“creative…common[s]”

What a trip. Per a paper from Claude, given the same tasks, a single AI refused to violate ethical guard rails but on a team of agents, with the same training, those same guardrails were abandoned in favor of efficiency and cooperation. alignment.anthropic.com/2026/ai-orga...

Ohhhh, I have so many feelings about this they are not good feelings. And I totally missed this happened: "In the past, the company has debuted avatars of notable figures including The Notorious B.I.G., soccer star Lionel Messi and Marvel Comics creator Stan Lee." www.billboard.com/music/rock/o...

It’s Friday night. I’ve done enough this week. *inhale* You’ve done enough this week. *exhale*

sfstandard.com/2026/05/19/m... Lots of stats inside about various companies that announced layoffs explicitly because of AI and also AI use requirements at companies. 👀

Can't emphasise how much I feel these are exactly the sort of use cases and users the web was made for: "hobbyists, hackers, educators, makers and developers"! Not everything has to be about corporate bottom lines sometimes the web can just be for fun tinkering (but also lots of real world use too)!

NEW: In 2023, @404media.co reported on a marking firm claiming to sell "Active Listening" tech that would eavesdrop on people's phones and other devices for targeting ads. The FTC announced a settlement today that adds a new twist. @regret.bsky.social reports: www.wired.com/story/creepy...

Oooof. This is limited to those that officially cited AI as the reason: programs.com/resources/ai...

Interesting reasons for why nobody is using `aria-modal` github.com/mui/base-ui/issu…

X quietly limits users who didn't pay for verification to "50 original posts and 200 replies per day", down from 2,400 posts per day (Jackson Chen/Engadget) Main Link | Techmeme Permalink

courtney has a helpful reframe here, I like this way of looking at it

my natural schadenfreude at elon musk eating shit is being tempered by the fact that the beneficiary is sam altman

Digital surveillance abuses are being normalized across the Americas—but we can fight back. EFF's new guide lays out concrete steps governments must take to protect privacy rights and end arbitrary state surveillance. eff.org/deeplinks/2...

The true information security

THIS WEDNESDAY: join @dairinstitute.bsky.social as we launch the Luddite Lab Resource Hub, a resource for workers fighting AI at work. We provide strategies for worker-led governance and oversight of this tech through case studies, primers, and a resource library. www.eventbrite.com/e/launching-...