Favorites: Record Protests, Coinbase Cameos, and Cosmic Dogfights

June 15, 2025: Doctors choose the wrong ChatGPT, SF forces interns underground, and households with two kids are sick half the year

Jun 16, 2025

See new posts Conversation The Lincoln Project @ProjectLincoln Trump parade vs No Kings march, San Diego — (src)(cached)

More people showed up for protests in all kinds of wild places than were at Trump's Birthday event. Also, unclear why there were Coinbase logos at Trump's parade (cached). A funny one: possibly more people protesting (cached) in El Cerrito, California (pop 28k) than at Trump's parade, although it's hard to get clear corroboration for that one. Size of the protest in Boise is wild (cached). This was the largest single-day protest in US history, larger than the Women's March (cached).

See new posts Conversation Visegrád 24 @visegrad24 Picture of an Israeli Arrow 3 interceptor carrying out an exoatmospheric hit in space, destroying an Iranian ballistic missiles around 100 km above earth. — (src)(cached)

Battles occurring in outer space do look pretty wild

See new posts Conversation Crémieux @cremieuxrecueil Companies are rapidly improving on GLP-1 weight loss. Regeneron just paired semaglutide (Ozempic/Wegovy) with a myostatin inhibitor and the result was... - More fat loss - Less muscle loss — (src)(cached)

In the thread there are also some details about how this will impact bodybuilding, although I'm not sure exactly what they're saying.

See new posts Conversation Rohan Paul @rohanpaul_ai A follow-up study on Apple's "Illusion of Thinking" Paper is published now. Shows the same models succeed once the format lets them give compressed answers, proving the earlier collapse was a measurement artifact. Token limits, not logic, froze the models. Collapse vanished once the puzzles fit the context window. So Models failed the rubric, not the reasoning. ⚙️ The Core Concepts Large Reasoning Models add chain-of-thought tokens and self-checks on top of standard language models. The Illusion of Thinking paper pushed them through four controlled puzzles, steadily raising complexity to track how accuracy and token use scale. The authors saw accuracy plunge to zero and reasoned that thinking itself had hit a hard limit. 📊 Puzzle-Driven Evaluation Tower of Hanoi forced models to print every move; River Crossing demanded safe boat trips under strict capacity. Because a solution for forty-plus moves already eats thousands of tokens, the move-by-move format made token budgets explode long before reasoning broke. 🔎 Why Collapse Appeared The comment paper pinpoints three test artifacts: token budgets were exceeded, evaluation scripts flagged deliberate truncation as failure, and some River Crossing instances were mathematically unsolvable yet still graded. Together these artifacts masqueraded as cognitive limits. ✅ Fixing the Test When researchers asked the same models to output a compact Lua function that generates the Hanoi solution, models solved fifteen-disk cases in under five thousand tokens with high accuracy, overturning the zero-score narrative. — (src)(cached)

I really enjoy this title.

Flotsam and Jetsam

– "A Democratic State Legislature passed a Voter ID requirement and the Republican Governor vetoed it." legislators are realizing things have changed, Voter ID laws are definitely not advantageous to Republicans anymore. (src)(cached)
– In the UK, the party of a lawmaker killed in political violence runs unopposed in a by-election, it sends a message that violence cannot alter the balance of power. Seems like a good idea! (src)(cached)
– Current state of the art in Geothermal drilling: 15,700 feet and 500°F at the bottom, drilled in 8 days. Last year it was 8000, 370°F in 41 days. This is huge progress! (src)(cached)
– Don't accept reasoning about current events that start from the premise that people hate Obama, or that they hated him at the end of his term. He was and remains the most popular living US President. (src)(cached)
– "Households with 2+ Kids Are Sick 56% of the Year" (src)
– Guest post on Silver Bulletin on problems with RCV voting occurring in the NYC Mayoral election. My understanding is that Approval voting still wouldn’t have this issue. (src)(cached)
– There are doctors using ChatGPT in professional contexts who are using the 4o model because 'it has the highest number' when the o3 model would clearly be superior. OpenAI's horrible naming choices might actually get someone killed! (src)(cached)
– Trump seems to have suddenly realized that going after illegal immigrants who are productive members of society might be a bad idea. Probably too much to hope that he now throws Miller under the bus (cached). (src)(cached)
– It’s illegal in SF to rent an apartment for less than a year, so summer interns are basically forced to rent illegally. Bad policy choices make people’s lives worse for no upside. (src)(cached)
– Somehow I had never heard this: carrots aren't good for rabbits, that meme is entirely because in Bugs Bunny’s first appearance, he’s making a reference to Clark Gable in It Happened One Night. (src)(cached)
– Trump is now saying they’ll try to tariff foreign films, which, don’t seem that popular? Seems to me that reciprocal tariffs for this would be extremely bad for the US. (src)(cached)
– Sadly, The Better Angels of Our Nature came out near the precise lowest rate of worldwide violence. Hopefully over the long term we’ll be able to get back on track and the last 10 years will be a blip. (src)(cached)
– Trump Is now humiliating the law firms that capitulated to his demands a few weeks ago. Plus Microsoft dropped their law firm who caved (cached). People want lawyers who will fight in court when they have the law on their side. (src)(cached)
– Trump has pardoned 700 years worth of prison time, mostly among January 6th rioters. (src)(cached)
– You probably already know about how tearing out waterfront highways has created beautiful waterfront space in cities like San Francisco. But I never wondered why they were built there in the first place. People weren’t stupid, it’s just that the area was full of industrial warehouses for the docks and no one could imagine space ever being desirable. (src)(cached)
– A minor note, but “The groups” is not just a made up trope. It’s how congressional staffers and other people in government talk and think about the progressive representatives who lobby them. (src)(cached)

Tugboat for Today

Discussion about this post