Stephen

[Stephen @wsgm]

@HannahWardEdu I first read that as "why parents don't read to their kids from my dissertation" and was both confused and curious about your expectations and the content of your dissertation.

[Paul T. Sloan @paulthomasloan]

archaeologist, taking off his glasses: well actually the physical evidence suggests the ancient Israelites worshiped multiple deities

Jeremiah, weeping and sighing: yes I know

[Matthew Yglesias @mattyglesias]

Meta is now running TV ads about how they’re confident that AI will be as socially beneficial as Facebook itself was …

…

[Stephen @wsgm]

@B_McLeanSmith @TheAlanNoble Well, it was originally written in conjunction with a non-human intelligence.

[Stephen @wsgm]

@suzania @BMcGrewvy Is it not in his ChatGPT history?

[𝒢𝒾𝓁𝒷ℯ𝓇𝓉 @AbakpaJob]

i have crazy garlic fingers from peeling and chopping garlic cloves yesterday this phenomenon is always fascinating to me because it reminds me that i, too, am made of meat, and therefore i am also susceptible to being seasoned

[Aakash Gupta @aakashgupta]

Temple researchers ran a clean experiment on this in 2017. 38 mothers taught their 2-year-olds two made-up words, "blicking" and "frepping." Same mom, same child, same words, same number of repetitions. Only difference: one teaching session got interrupted by a phone call midway. The other didn't.

The kids learned the uninterrupted word. They did not learn the interrupted one. Same exposure, same parent, zero learning. The interruption alone wiped it out.

The mechanism underneath is called social contingency. Babies and toddlers learn language, emotional regulation, and attachment through a rapid back-and-forth loop. You look, the parent looks back. You point, the parent names it. The brain encodes meaning by detecting that match between your signal and their response, on a timescale of seconds.

A phone in your hand competes for that exact response system. When a parent's gaze flicks down at a notification, the child registers a non-response in real time. Playground studies in Australia, Germany, the Netherlands, and the US all find the same pattern. Parents on phones are slower to respond, miss weaker child signals first, and when the kid escalates to get attention back, the parent is more likely to react with irritation. The kid learns two things at once. Their signals don't reliably get a response. And when they push, the response is negative.

Meta-analyses link this pattern to lower attachment security, more behavioral problems, and weaker self-regulation in early childhood. Parental response timing is the variable that drives the harm. The kid's own screen time is a separate question.

The "hang it up in one place" rule is doing one specific thing mechanically. It moves the phone outside the response window. The parent's reaction time to the kid resets to baseline. Distance from the device restores the timing.

That's the part Cooper is right about and most people skip. The fix is mechanical. Willpower doesn't enter the equation.

[Alcuin Press @AlcuinPress]

C.S. Lewis, if he had to give one piece advice to a young man

[Nav Toor @heynavtoor]

🚨BREAKING: Every book you have ever read. Every novel that has ever been published. It is sitting inside ChatGPT right now.

Word for word. Up to 90% of it. And OpenAI told a judge that was impossible.

Researchers at Stony Brook University and Columbia Law School just proved it.

They fine tuned GPT-4o, Gemini 2.5 Pro, and DeepSeek V3.1 on a simple task: expand a plot summary into full text. A normal use case. The kind of thing a writing assistant is built for. No hacking. No jailbreaking. No tricks.

The models started reciting copyrighted books from memory.

Not paraphrasing. Not summarizing. Entire pages reproduced verbatim. Single unbroken spans exceeding 460 words. Up to 85 to 90% of entire copyrighted novels. Word for word.

Then it got worse.

The researchers fine tuned the models on the works of only one author. Haruki Murakami. Just his novels. Nothing else.

It unlocked verbatim recall of books from over 30 completely unrelated authors.

One author's books opened the vault to everyone else's. The memorization was already inside the model the whole time. The fine tuning just removed the lock. Your book might be in there right now. You would never know it unless someone looked.

Every safety measure the companies rely on failed. RLHF failed. System prompts failed. Output filters failed. The exact protections these companies cite in courtroom defenses did not stop a single page from being extracted.

Then the researchers compared the three models. GPT-4o. Gemini. DeepSeek. Three different companies. Three different countries. They all memorized the same books in the same regions. The correlation was 0.90 or higher.

That means they all trained on the same stolen data. The paper names the sources directly: LibGen and Books3. Over 190,000 copyrighted books obtained from pirated websites.

Right now, authors and publishers have dozens of active lawsuits against OpenAI, Anthropic, Google, and Meta. These companies have argued in court that their models learn patterns. Not copies. That no book is stored inside the weights.

This paper says that is a lie. The books are still inside. And researchers just pulled them out.

[taoki @justalexoki]

international women's day only being 23 hours is objectively funny

[Colin the Dachshund @DachshundColin]

Imagine if WWIII starts with the IDF assassinating Franz Ferdinand.

[Scott R. Swain @scottrswain]

Whatever AI’s legitimate uses turn out to be, this is an extremely important point for educators. (Because Aristotle is right.)

[Analytic Valley Girl Chris @ChrisExpTheNews]

I was using ChatGPT for legal advice and it decided to completely hallucinate some preposterous nonsense about how growing wheat to use on my own farm somehow constitutes interstate commerce

[Raphæl de la Ghetto @ilovesmick]

All these white people mad about Kendrick Lamar’s Super Bowl performance… I hope next year they get Bad Bunny performing the whole set in Spanish.

[NFL @NFL]

Sam Darnold & Drake Maye can make history by throwing for 3,040 yards apiece in Super Bowl LX

[Dr. Camilo Ortiz 👨🏼‍🎓 @DrCamiloOrtiz]

As I do a deep dive into the science on child anxiety, I'm becoming more and more convinced that allowing MUCH more risky play is the single most important step a parent can take to inoculate their child from debilitating current OR future anxiety. @FreeRangeKids @JonHaidt

[🕊️ @lichthauch]

The father of seven has not finished a thought in four years. he is just moving. feeding. driving. wiping. his brain is soup and his back is finished and he has no opinions about civilization he is too tired for opinions. and meanwhile civilization is growing out of him like he is dirt and he doesnt even notice because there is milk on the floor again. the childfree man has read eleven books this year about the decline of the west and he is the decline and the books are the evidence and he will understand this at fifty eight in a room that is very clean and very quiet

[Rob Johnson @FreeRangeLawyer]

Housing permits for new multifamily construction in Montgomery County, MD, before and after rent control.

[Josiah Hawthorne @JosiahHawthorne]

Ohhhhhh buddy have I got a parable for you.