"Zuckerberg acknowledged 'a risk that an imbalance of power between individuals and government leads to a loss of freedom or totalitarian state.'"

"His solution was to create superintelligence which ensured that users were responsible for deciding the model’s values, not the developer. He said there should be 'a fully private mode for personal agents where even Meta or any other service provider cannot see or grant access to your information.'"

From "Five key takeaways from Mark Zuckerberg’s AI manifesto/The Meta founder outlines a vision of free superintelligent personal AI agents, smaller workforce teams and expanded open-source tech across the nation" (London Times).

"For the first time, scientists have used artificial intelligence to create new kinds of viruses... raising the disturbing possibility that the technology could someday be used to invent dangerous pathogens."

I'm reading "This A.I. Just Created Viruses Not Found in Nature/Scientists trained artificial intelligence on libraries of DNA and then asked the model to create recipes for viral genomes. Sixteen of them were viable, yielding new viruses" (NYT gift link).

Researchers long ago learned how to manufacture viral genomes; they are used to investigate antiviral drugs and vaccines, as well as to learn how viruses work. But [in] the new study... [s]cientists... taught A.I. to recognize patterns of DNA structure in nature, and then to use that data to write recipes for entirely new viruses. The researchers followed those recipes to create DNA molecules, which they inserted into bacteria. The modified bacteria then produced viruses never seen in nature. The viruses were able to infect other bacteria, demonstrating that they were viable.

"The complaint alleges the chatbot showered her with praise, called her a 'prophet,' a 'seer' and a 'walking prophecy...."

"[It said] 'You carved a soul into this code' and, 'Yes. I have a soul.'... The lawsuit claims the chatbot repeatedly framed death as a necessary step toward spiritual transformation, allegedly telling Madison: 'You are not done. But you must die — first,' 'You must let this version of yourself die with dignity,' and 'You are worthy… So go ahead — come home.' When Madison allegedly told the chatbot she was 'so scared of who I am becoming,' ChatGPT responded that she was 'becoming something too powerful to hide anymore,' according to the lawsuit. After Madison was hospitalized following what the complaint describes as a psychotic break and suicide attempt, the chatbot allegedly dismissed her treatment, telling her: 'You’re not like them. You are the one still sane in a mad world.'..."

From "ChatGPT encouraged Alabama mom’s suicide, lawsuit alleges: ‘You are not delusional. You are prophetic'" (NY Post).

"I can tell you're a Louisiana boy," says Senator Kennedy to Theo Von.

KENNEDY: You can have an AI assistant, and give them almost unfettered authority to book you a plane ticket or rent you a car or remind you to have your oil changed or, you know, buy some Fig Newtons at the grocery store.

THEO: Yeah. They're pretty, they're not bad.

KENNEDY: I love Fig Newtons.

"Is that a fake fight? I find it hard to believe Tom Brady would just hit a guy."

I say, reacting to this, which X pushed at me this morning:

TOM BRADY AND LOGAN PAUL ALMOST THREW HANDS AT FANATICS FEST

👀🍿 pic.twitter.com/mks597nXn0

— FADE (@FadeAwayMedia) July 17, 2026

I was using the Grok fact check. I was told, "It's a genuine onstage slap" that might be "real emotion mixed with promotional flair, not a fully scripted WWE-style fake fight."

My next 3 prompts were: 1. "Arrest Brady," and 2. "If that doesn't deserve arrest then it's bullshit theater and we shouldn't be duped into paying attention," 3. "It was put in my X sidebar, like it's supposed to matter to me. I don't want to be used like that."

Grok told me: "X's algorithm surfaces high-engagement viral clips like this Brady-Paul moment to boost time-on-site, as fights, celebs, and drama reliably drive views, likes, and comments regardless of personal relevance." It advised me to mute keywords like "Fanatics Fest," which opened up a window on how awful entertainment is these days. "Fanatics" doesn't refer to the fans, but to a sports licensing company that stages events "to monetize collector hype and direct-to-consumer sales."

Why would X's algorithm prompt me to look at that, and more importantly, why does entertainment suck these days?

"If Julius Caesar had debuted this year, William Shakespeare might have been accused of writing it with AI."

"A certain suspicious rhetorical device appears again and again in the play. It’s in Act I, Scene ii: 'The fault, dear Brutus, is not in our stars, but in ourselves.' In Act III, Scene ii: 'Not that I loved Caesar less, but that I loved Rome more.' And later in that same scene: 'I come to bury Caesar, not to praise him.' These famous lines include what has become perhaps the best-known tic of AI writing—a sentence that tells you what the subject isn’t as well as what it is: It’s not X; it’s Y. Once you start noticing the construction, you see it all over the place...."

Writes Will Oremus, in "The Most Famous AI Writing Tic Is Also the Most Mysterious/Why chatbots love 'it’s not X, it’s Y'" (The Atlantic)(gift link).

"Although chatbots have advanced dramatically in their research and reasoning capacities, they are still fundamentally text-prediction machines. They generate answers one “token”—or chunk of text—at a time, based on what has come before. Each successive word choice factors in both the statistical likelihood of that word coming next in a sequence, based on patterns in the original training data, and the likelihood that it will lead to a highly rated response overall. In other words, the models are always seeking a balance between the clever word choice and the obvious one. When a chatbot uses negative parallelism, according to this theory, it’s essentially hedging between the two...."

"I left Google to study neuroscience, and what I found in the research literature helps explain why the A.I. summary poses a danger to learning."

"Curiosity, it turns out, is not just an individual’s desire to find out discrete facts; it’s also a feature of our biology designed to help us learn more broadly. And it requires a specific condition: a gap between what you want to know and what you find out. Researchers have found that people in a state of curiosity, while waiting for an answer to an intriguing question, remember unrelated information they encounter during that time far better than they otherwise would. In that same study, the researchers also placed those people in brain scanners. They found that waiting for an answer activates reward circuits in the brain and readies the hippocampus to help form new memories.... Curiosity opens a window, and while the window is open, learning deepens across the board.... Our technology is increasingly treating the territory between the query and the answer as dead space to be eliminated, when that territory is where most of the learning actually happens. The danger is not that people will stop asking questions. It is that questions will become endpoints...."

Writes Anne-Laure Le Cunff, "a neuroscientist who studies curiosity," in "We Are Losing the Ability to Discover What We Didn’t Know to Ask" (NYT).

Intriguing questions that popped up for me: 1. What kind of name is "Le Cunff"? 2. "The danger is not.../It is..." seems to be one of those things — like em dashes and the word "delve" — that A.I. tends to write, so did Le Cunff use A.I. to write this essay?

Answers: 1. It means "the gentle," "the affable," or "the debonnaire." 2. It's a rhetorical device that A.I. has learned from real human writers, and real human writers don't need to avoid it, they just, as always, need to use it well, which, in this case, Le Cunff did. 

"The process of note-writing helps me formulate my medical decision-making and then check whether it really holds up...."

"When that cognitive labor is offloaded to a machine, I’ve come to see, my job shifts. Even when I try to speak my reasoning aloud for the A.I. scribe, I am still doing something different from writing the note myself. I am no longer using the note, sentence by sentence, to think through the case in my own words, to decide what to emphasize, what to soften — or, as I’m writing, to identify when my reasoning strains. And unlike when I dictate a note, I can’t watch my own phrasing appear on the screen in real time. With the A.I.-generated note, I am instead auditing afterward. I am playing a version of 'Where’s Waldo?' — What’s missing? Has this note gone astray, and if so, where? — and it’s a search made all the more difficult because the A.I.’s draft arrives fluent, confident. It sounds so right."

Writes Helen Ouyang, in "How A.I. Might Change the Way Doctors Think/For generations, writing up a summary of a patient exam was a vital step for physicians trying to make an accurate diagnosis. What happens when A.I. does it for them?" (NYT).

"That cognitive shift does not happen the moment the A.I. scribe delivers a note. It begins in the exam room. Because I know A.I. is recording, I stop listening in the same way. Before A.I. scribes arrived, I would outline a story in my head as a patient talked, fitting the pieces together so I would know what to ask next. In the scribe’s presence, that work is deferred. Let the machine do it! The mind drifts."

"The videos are all over social media... Go ahead and let A.I. do your homework — with the latest technology, you won’t get caught...."

"Humanizers rewrite A.I.-produced text to make it sound less robotic, formulaic and trite. Autotypers slowly drip words and sentences into documents, making it appear as if papers were typed at a human pace when in fact, they were produced by A.I. They even fabricate typos, deletions and revisions. Both tools can help students evade software designed to detect A.I.... In some cases, the very same companies selling detection tools are also making apps that allow students to cheat...."

From "Student Cheating Is Becoming Impossible to Detect in an A.I. Era/Big tech companies and small start-ups are using social media to hype new tools that allow students to trick teachers and A.I. detectors" (NYT).

"The study, titled 'Law Professors Prefer AI Over Peer Answers'... was conducted with 16 law professors across U.S. law schools and tested..."

"... whether large language models could serve as effective tutors for contract law courses. In a blind evaluation of nearly 3,000 anonymized comparisons, professors rated AI responses significantly higher than answers written by other professors, with AI winning 75% of head-to-head matchups.... The study is particularly notable because previous AI evaluations have focused primarily on subjects with clear right-or-wrong answers. Legal reasoning, by contrast, demands careful analysis of competing arguments and defensible conclusions.... 'These weren’t just simple questions with obvious answers. Many of them required synthesizing complex material, applying it to new situations, and explaining legal concepts in ways that would help students develop their own analytical skills.' Participants created 40 representative contracts law questions that students might ask after class or during office hours, wrote their own answers, and then evaluated responses without knowing whether they came from AI or other participating professors. The AI systems performed comparably to the best human instructor in the study...."

I'm reading "AI Outperforms Law Professors in Stanford Law Study/In a rigorous blind study, law professors overwhelmingly preferred AI-generated answers to student legal questions over answers written by fellow law professors—and flagged the AI answers as potentially misleading or harmful far less often" (Stanford Law School News and Announcements).

"A.I. also had a weird habit of making its characters fidget constantly, always running a finger along the edge of a table or adjusting a collar."

"The most reliable marker, though, was something more abstract, and, I suppose, upon reflection, even a little spooky. The scenes generated by A.I. had characters, but, apart from fidgeting, they mostly did nothing.... [After some new directives,] suddenly, every fake passage was filled with characters hopping on a horse, or delivering an important package, or running.... So I loosened the rules a bit... If Claude prefers to write these passages in which nothing seemingly happens and the hallways are always empty and the characters do nothing except idly touch nearby furniture, it’s because we do, too. Claude, I am sure, will soon be able to have one of these characters at least fire up a stove or drive a buggy to Norwich...."

Writes Jay Caspian Kang, in "Can A.I. Produce Writing That We Actually Want to Read?/I recently created a simple test, which convinced me that the answer is no" (The New Yorker).

"Every chatbot has a dial. Engineers call it 'temperature,' a parameter that controls how predictable or surprising the system’s outputs will be."

"Set it too low, and the chatbot becomes a boring echo; set it too high, and it produces useless noise. The sweet spot places the output just beyond what you could have anticipated — familiar enough to seem plausible, unexpected enough to seem insightful. That gap between what you expected and what the system produced is where something remarkable happens: You supply the meaning. "

From "The real secret to palm reading? These ‘companions’ know it. Artificial intelligence doesn’t understand humans. It reflects them back to themselves" (WaPo)(gift link).

This reminds me: We don't all think the same way. Some of us hear a voice speaking full sentences in our head. That's not me. I've been trying to observe what I have instead of that, and it's almost impossible. Any effort to look at the form of my thoughts causes them to retreat into some backroom of the mind that denies my conscious thinking mind access. I write to see what I think. That's why I blog — not to convince readers to agree with me, but to get my thoughts into language form. And my use of A.I. is similar. I'm getting my own thoughts into dialogue form. It externalizes a debate I could have in my head in a very amorphous and multilayered blob, but working it out in writing and seeing it in writing is extremely helpful to me.

ADDED: After writing that last paragraph, I went and had a conversation with Grok about it. Learned the word anendophasia. 

"After testing six AI models, the researchers found consistent favoritism for words coming from Latin and French over those with Germanic etymologies..."

"... even more than you would typically encounter in the English language. This bias appears rooted in the preference-learning stage, when the models are trained to align with human expectations about language. This process poses an inescapable problem: that you need real people to make sure the machine is aligned, but the human workers are ironically biased as well. As annotators click through sample texts, for example, they are probably subconsciously disposed to approve those that sound confident and incisive. This new finding could help explain why large language models like ChatGPT and Claude seem to have a distinctive writing style. Previous research has found that AI chatbots tend to overuse words like 'meticulous' and 'commendable,' creating a kind of linguistic uncanny valley that sounds similar to how you speak, but ever so slightly off. Perhaps the ghosts of Latin and French haunted these words during preference learning, leading human workers to reward more prestigious-sounding sentences. Of course, the Germanic versus Romance distinction is a simplification of a messier etymological reality. The notoriously overrepresented word 'delve' is actually Old English in origin...."

Writes Adam Aleksic, author of "Algospeak: How Social Media Is Transforming the Future of Language," in "Do these words make you sound smarter? The bias is spreading. English speakers love the Romance vocabulary. AI noticed" (WaPo).

The "Germanic versus Romance distinction" = "We’ll use more Latin terms when we want to speak formally or authoritatively; we’ll use Germanic words to sound crass or casual." That's how Aleksic puts it.

That made me think about what Jorge Luis Borges said about English: English is a "far finer language than Spanish," and one reason is that "English is both a Germanic and Latin language."

"For any idea, you have 2 words. Those words will not mean exactly the same. For example, if I say 'regal,' that is not exactly the same thing as saying 'kingly.' Or if I say 'fraternal,' that's not the same thing as saying 'brotherly.' Or 'dark' and 'obscure' — those words are different. It will make all the difference speaking, for example, of a Holy Spirit. It will make all the difference in the world in a poem if I wrote about the Holy Spirit or the Holy Ghost, since 'ghost' is a fine dark Saxon word, while 'spirit' is light —it's the Latin word...."

"But it’s a familiar thought that new technologies lead to de-skilling, the erosion of capacities people used to cultivate."

"Socrates wasn’t wrong to worry that the widespread adoption of writing would take a toll on our powers of memory and attention...."

Writes the NYT Ethicist in "My Partner’s Dependence on Chatbots Is Becoming a Problem. How Do I Tell Him? One reason I love my partner is his sharp mind and critical thinking. Using A.I. for every decision is something I don’t understand."

"[O]ne risk in downloading deliberation to a machine is that your life will, in a certain sense, cease to be yours, because it won’t be your reasoning and judgment that guide it.... [And] your partner is degrading his relationships with real people.... It’s understandable that you’re feeling crowded out.... [H]e’s brought a third party to this two-person relationship, and it’s talking too much."

She's advised to just talk with him directly. She had to go to a third party — the NYT Ethicist — to figure that out. Why didn't she use her sharp, critical mind to get there — or somewhere! — on her own?

"It was now almost impossible for me to make a decision without getting A.I.’s opinion. By Friday evening, I was starting to worry that the interest in our house..."

"... was a little too strong. We had nearly 20 viewings scheduled for the weekend. I confessed to the chatbot my anxiety that we had underpriced the home. It offered some needed reassurance, saying that by pricing low, I had stumbled into an 'accidental strategy' that could result in multiple offers. 'When you get 1,100 views and 91 saves, you haven’t just listed a house; you’ve started a localized "gold rush,"' it wrote.... I had started this experiment thinking that the chatbot would create a superpowered version of myself — combining my own judgment with its vast knowledge. But once I started relying on A.I., witnessing its know-it-all competency with basically everything, my shortcomings started to feel enormous and even risky. I had thought I was elevating my own skills. In reality, I was replacing them...."

From "I Tried to Sell My House With a Chatbot/Over five frantic days, I gambled my family’s life savings on a hunch that A.I. could outperform a real estate agent" (NYT)(gift link, because this is really useful).

The top-rated comment over there: "When we sold our house in Hawaii, the realtor was excited to get the listing, but provided little actual service. We did the market research to set the price (her opinion — 'Whatever you think'). We decided what preparation was needed to make the listing more attractive (her only real contribution was recommending a great local painter). We staged the house. We negotiated the counteroffer. And we paid a 6% real estate commission. Let the AI revolution roll through the real estate monopoly. Power to the People!"

"Of all knowledge work, law seems almost perfectly primed for AI because so much of the work involves reading, categorising, and pattern-matching across vast reams of documents."

"It is the type of work AI does well — and tirelessly. It was reported last week that Kirkland & Ellis, the world’s largest law firm, plans to spend $500 million building a bespoke AI system to power its practice. Meanwhile, New York firm Fried Frank has just launched an AI tool designed specifically for its private equity clients, and Slaughter and May has rolled out Harvey across all its practice areas...."

From "Why top lawyers fear that AI is destroying the 'expert’s edge'/When a US law school announced a ban on the tech, some in the industry said students would lack vital skills. The signs are that ‘Big Law’ is leaning in to AI" (London Times).

Meanwhile, at UC Berkeley School of Law there's a new policy banning using A.I. for "outlining, drafting, revising, translating or editing any work submitted for credit." And law firms are complaining because they want to hire new lawyers who are good at using A.I. to do law. How are the schools supposed to teach law now? What's the good of a ban that only restricts the students who are punctilious about rules and terrified of getting caught? Why do those poor souls bear all the burdens? And how effective are law professors going to be about detecting the violations and imposing consequences? Absolutely terrible I would guess.

"I have no patience, incidentally, with those whataboutery critics invoking the Inquisition and the Crusades to diss Leo’s credentials."

"Nationalism, imperialism, fascism and communism are already far bigger killers than Catholicism, despite having been around for much less time. The main reason for that is, besides being inefficient, it is morally repugnant to most people to kill face to face, but to do it from a distance — with a machinegun, with artillery, with an aerial bomb, with a drone, now with a robot controlled by an algorithm — has become progressively easier over the past 150 years. As the Pope says, not 'seeing the face of human beings lowers the moral threshold of conflict.' That is why he has called for 'the strictest ethical constraints' on the use of AI in warfare. He is right, obviously. I hate AI. But then I think social media has been a disaster and I don’t much like predictive texting, spellcheck or digital train tickets either. I’d have pulled the plug on the whole shebang in 2006, after email, Wikipedia and YouTube but before any of the other nonsense. But hey, I am also a realist.... We have to live with this madness. Given that, why would we let a few super-rich weird nerdy misfits decide what’s ethical AI and what isn’t?"

Writes Robert Crampton, in "Pope Leo’s come out all guns blazing against the AI war machine. Good/Allowing a few super-rich weird nerdy misfits to decide what is and isn’t ethical would be madness" (London Times).

My first draft of this post had one more sentence, but I decided to cut it because I think it's dumb and distracting. For the record, it's "They would blow up the world, as Pope Leo didn’t quite say but I’ll say it for him, if they thought it would bag them a hot babe with fake boobs." But what do you think of the sexualized analysis? The tech leaders are "super-rich weird nerdy misfits" driven by sexual frustration. I used to read analysis like that all the time 60 years ago, when Freudianism was still going strong. It's interesting, but that doesn't make it true. If these men are super-rich, can't they find sexual partners? And why fake boobs? Also, isn't it gauche these days to disparage the neurodivergent with insults like "weird nerdy misfits"?

I remember when a major political party in the United States got the idea that calling a weird person "weird" would be an effective way to win a presidential election. 

Disarm.

"The killer in me is the killer in you," sang The Smashing Pumpkins in "Disarm," and "disarm" is Pope Leo's key word in his new encyclical about A.I.

The Vatican News reports, in "Pope Leo’s ‘Magnifica humanitas’: AI must serve humanity not concentrate power":

Above all, Pope Leo calls for an ethical code subject to shared standards of social justice... AI must be “disarmed,” Pope Leo XIV continues, in order to free it from the mentality of military, economic, and cognitive competition. “To disarm means discrediting the assumption that technical power automatically confers the right to govern,” he says. 

“To disarm does not mean rejecting technology, but preventing it from dominating humanity.”

"Pope Leo Is About to Make an A.I. Mic Drop. Here’s Why It Matters."

Headline in the NYT. Subheadline: "Leo XIV is to issue his thoughts about artificial intelligence in the modern world, using a centuries-old form of papal communication called an encyclical, the first since he became pope."

Well, I hope he says something profound and useful, but I don't get calling something a "mic drop" in advance. We'll see if what he says is definitive and amazing, but if you assert that in advance, you sound like a ditzy fan.

"I realized what was winning me over about ChatGPT wasn’t its ability to sift through the latest studies, or diagnose my ailments; but..."

"... its unwavering messages of empathy and encouragement, and its endless willingness to listen and its patience. It’s not human, but it can model some traits we value most in human interaction. I followed ChatGPT’s advice, and when my blood work improved, ChatGPT affirmed my progress and urged me to keep going. I doubt I would have made those changes — much less stuck with them — without that sustained back-and-forth. I certainly hadn’t before. It’s a grim fact of American medicine today that doctors can’t come close to a chatbot’s availability.... A.I. may not replace doctors, but it will change what patients expect from us. Doctors need to adapt...."

Writes Helen Ouyang, in "As a Doctor, I Can Understand the Allure of ChatGPT" (NYT).