If you want to "just do biology and never worry about programming" the answer looks like adequate funding for research software engineering so we can make validated, shared tools that reduce the need for new code to none. Not everyone generating analysis on demand that is impossible to validate.

MadScientist@mads100tist.bsky.social · last wk.

It is asinine to say you need to use an opaque algorithm you have no control of to do something you were absolutely able to do three years ago. And that's before we even consider theft and ethical implications

Who would have thought that the radical project of doing things that help people and make the city work better would be the thing that makes people like you and probably vote for you again.

jerome cylin@jeromecylin.bsky.social · 2w ago

Actual small government: "Bodega owners will no longer need a separate license to sell the same products outside the store like fruit and flowers – as inside. Restaurants will no longer need a “superfluous” frozen dessert permit to sell ice-cream on top of the permit already required to sell food."

Actual small government: "Bodega owners will no longer need a separate license to sell the same products outside the store like fruit and flowers – as inside. Restaurants will no longer need a “superfluous” frozen dessert permit to sell ice-cream on top of the permit already required to sell food."

I thought this was a pretty interesting take from Codeberg's announcement of their anti-LLM policy They argue that many developers using LLMs are essentially heaven-banning themselves where they don't realize they're developing for a nonexistent community

A reminder that UC _faculty_ and Faculty Associations — not the universities themselves nor their administrators — are waging this battle in court, demonstrating great bravery and moral clarity.

White House admits it used keywords to kill billions worth of California research grants

Federal agencies that terminated more than a thousand research grants at the University of California last year admitted that they used keywords related to diversity, gender, vaccine hesitancy and COV...

apnews.com

Idk maybe that nuance is important to me because you can work with someone who has misinformed views about AI and bias. You can’t work with someone who knows AI can be racist and they just don’t want to take responsibility for it

People view AI as “unbiased” in the same way that people have less issue with statistical versus taste-based discrimination. As long as you can ground it in numbers by saying people with x characteristics are more likely to do y. And never ask about the training data.

The Breach@breachmedia.ca · 3w ago

Ontario jails are using an AI tool that claims to predict prisoners’ behaviour. And it’s disproportionately placing Black prisoners in maximum security. An investigation by @desmondcole.bsky.social reveals the province knew the tool would discriminate against Black people—and used it anyway.

As a robotics researcher, learning scientist, and author who has followed Realbotix for years, I'm compelled to do a Bluesky thread on this whole pivot that Realbotix is now trying, turning from humanoid sex robots to classroom "education" robots. Buckle up. nysfocus.com/2026/07/14/n... 🧵

Rural NY School To Launch Humanoid Robot Teacher

Plans to deploy a "lifelike" robot and avatar teaching assistant at Salamanca High School have sparked debate over the role of technology in schools.

nysfocus.com

We (faculty at USC) took on an institutional behemoth with all the money and clout on earth, and we WON. We won our union. We're the biggest private-university faculty union in US history. Trying to make people care about this, challenge level IMPOSSIBLE

The revelations came out in a federal court case brought by Democratic states challenging ICE's access to Medicaid data to aid in deportation efforts. n.pr/4f9OdMn

ICE shared Medicaid data it wasn't supposed to have with Palantir

The revelations came out in a federal court case brought by Democratic states challenging ICE's access to Medicaid data to aid in deportation efforts.

n.pr

we've been doing this too but more as performance art to see what we can get the bots to do. so far the answer is "anything" github.com/dwebagents/A... i think my favorite is the PR whose only text was "butter" which got us the butter mode page: dwebagents.github.io/AgentPipe/bu...

Andras Bacsai@heyandras.dev · 3mo ago

We made a fake repo with fake bounties, and the bots are applying fake PRs, so we know who is fake, and we can ban them from the Coolify repo. IQ over 1000

With two people dead in Memphis in four days and one the same week in Houston, that's four people federal agents have shot and killed over just nine days. July 5: Tyrin Johnson July 7: Lorenzo Salgado Araujo July 8: Alfonso Ivy July 13: name not released yet

Bloomberg News@bloomberg.com · 3w ago

A person was killed in a shooting involving US Immigration and Customs Enforcement in Maine on Monday morning, the speaker of the state’s House of Representatives said

here is a chess variant where the DVD logo is a semi-playable NPC that bounces around the board. complete with stockfish fork so the cpu can actually understand DVD and realistic money sound effects jon-e.net/dvd-chess/

dvd chess

listen up everybody what if chess but it had dvd as a semi-playable npc where dvd bounces around and captures it all

jon-e.net

Scoop: NIH is on a mission to enlist one million Americans to share their genetic data to help with health research through a precision medicine program called “All of Us.” What participants don’t know is that at one research site, Palantir stores the data. www.motherjones.com/politics/202...

Palantir has a hand in NIH's most ambitious health initiative

The defense tech firm's work with public health data has largely flown under the radar.

motherjones.com

#followthemoney www.struggle-la-lucha.org/2026/07/03/1...

10,000 ICE arrests in five days — Wall Street demanded more detention

Police assault a protester outside Delaney Hall in Newark, New Jersey, where demonstrators have demanded the shutdown of the GEO-run ICE jail. Outside Delaney H

struggle-la-lucha.org

ITS THAT TIME AGAIN

great piece, amazing how the silicon valley ideological space can make the transition from "self evaluating your behavior and thoughts" to "literal harry potter magic and telepathy" seem almost inevitable.

robustly reproducible if you've ever tried to design prompt injection or worked on related but nonoverlapping projects - the context window of an LLM structurally invalidates it as a scientific instrument because it is structurally opaque, because it is a consumer product, not a scientific one.

Josh Pasek@joshpasek.com · last mo.

Just inadvertently discovered that a whole bunch of research on AI responses may be invalid. Apparently, even with memory off, ChatGPT is still giving contextual information, including the name of the user! It used my full name casually in conversation; which means that data is in the model.

Yooo

NEW: Insane screwup inside Meta. The company exposed worker keystroke data that was being used to train AI — making it potentially accessible to anyone at the company. The data included personnel and performance info, private convos, full transcriptions…imagine your coworkers seeing all of that.

Meta Exposed Data Internally From Its Controversial Employee-Tracking Program

Employees had previously raised concerns about the initiative, which involves collecting workers’ keystroke data to train AI models.

wired.com

have reviewed a few AI citation correction packages, and they would all amplify the problem - if invented male authors are present, hit crossref, pick the first match. Authors are more exact matches than titles. "Better LLM tooling" does not fix this, you can't dig up.

Mel Andrews@bayesianboy.bsky.social · 2mo ago

When utilized in literature review, LLMs consistently 1. fail to mention female authors in female-led literatures, 2. insist that men are more influential or more heavily cited when this is contradicted by objective citation counts, and 3. attribute women’s work to hallucinated male scholars.

What they teach you in business school is that its good to make your employees feel as if they are trapped in a senseless universe where they have no control and their every action could be undone by a random number generator that cannot be turned off.

Michael Okun@michael-okun.bsky.social · 2mo ago

I emailed and met with the editorial office to ask for the AI assistant to be turned off. I was told this is not possible. Instead, I was treated to some vague promises of potential future improvements and a dose of gaslighting.

I’ve officially resigned as Associate Editor for Frontiers in Systems Neuroscience. It used to be a reputable journal, but became a case study in how forced automation destroys academic integrity. 👇

Zotero files for its IPO, chasing exit liquidity after raising a trillion dollars on a product with only a few billion in revenue, hoping it has hooked enough academics on formatted citations to move to a per-cite billing model at 100x the price

I think everyone who gets subjected to AI nonsense yearns for the "fuck off you're banned" button

Jason Koebler@jasonkoebler.bsky.social · 2mo ago

In Mississippi, lawyers on both sides of a court case were caught using AI, essentially two LLMs arguing against each other. The judge dismissed all lawyers from the case and canceled the trial: www.404media.co/judge-learns...

Asking again, what are the "human looking vaguely liberal repost engagement bait" filters/labelers people are using nowadays?

I knew someone who was a freelancer for chumbox travel blogs long before AI, and what they did is basically look up place names on google maps and say it was good or bad pretty much at random.

Gershom Gorenberg גרשום גורנברג@gershomg.bsky.social · 2mo ago

Reporter for The Free Press says there’s no point in going places yourself if you can read about them on Wikipedia. Truly, the soul of a journalist.

Many builders in the Open Social Web are discouraged by the lack of funding in the ecosystem, slow pace of feature development (e.g. private accounts), and “low” adoption relative to incumbents like Meta. In my keynote I emphasized multi-generational struggle & movement building. People who faced…

for "AI" to be profitable, costs need to be aligned with token use rather than subscription, and then everyone needs to use orders of magnitude more tokens. "Agentic" workflows are for making massive token spend palatable by making the point of comparison "20 engineers" despite being the same thing.

Eryk Salvaggio@eryk.bsky.social · 2mo ago

This image from Anthropic goes out to the two guys on here who called me dumb for talking about stochastic flocks

🚨BREAKING: ICE Protesters clash with police on Day 13 of Delaney Hall ICE Prison hunger strike protests. "You're doing the bidding of ICE! You're working for a concentration camp! You're f*cking Hitler! WE'RE LIVE ⬇️