Joseph Galarneau

Just for context, 68 quadrillion miles cited in @nytimes article is 11,570 light years or 10% the diameter of Milky Way galaxy or halfway between our location and the massive black hole at the center of our galaxy. It just seems off to me. nytimes.com/2026/06/11/sci…

And another Central Park view while running at sunset, this time looking to the east over the reservoir.

@skalskip92 But SAM3 consumes a crazy amount of resources to do that.

there's no catch; SAM3 is open source and really good one of the things it does really well is object tracking, even in crazy complex scenes like basketball probably my favorite computer vision model ever

Had to stop during a recent run in Central Park for a quick shot.

Alan Greenspan -- happy belated 100th birthday!

Lorne Michaels: “The thing that kills most things is overthinking.” nytimes.com/2026/04/22/mov…

Been @nyrr member off and on for more than two decades, but no longer see the point. Races sell out in minutes months in advance. And costs gone through the roof. Maybe if I’m training for marathon, but otherwise, nope.

This can only mean one thing: Allergy season in NYC.

Doom loop at Reddit? One of biggest sources for GenAI training but now its bots reject majority of posts because they suspect (sometimes wrongly) that they’re AI generated. Trying to avoid slop creating slop but now losing a lot of good content & frustrating users. Can’t win.

AI + creative data + product marketing: Join my NYC team! jobs.ashbyhq.com/Vidmob/fb9010c…

And it's only April. Who else is feeling this way?

The rise of synthetic mediocrity in AI ad creative. It doesn't have to be that way. vidblog.vidmob.com/blog/distincti…

Now 100 miles from home after 25,000 miles & 46 hours in the air on 7 flights. All on time or ahead of schedule but thanks to traffic into JFK, going round in circles over Scranton. Only flown 20 hours today so what’s another 60 minutes?

Happy that @JohnHMcWhorter has returned to Lexicon Valley. I, for one, appreciate doses of linguistic wonkiness sprinkled with show tunes.

Is there a department at Apple that takes perfectly good products and renders them less useful in future iterations? Like AirPod form factor that use to sit comfortably, almost effortlessly, in your ear and now that it's bigger in AirPod 4, keeps wanting to slip out.

@aakashgupta Wrapper company using a standard IDE. No product mojo that can't be easily knocked off -- it's just distribution. The past few months have shown us this can be a shallow moat in AI land.

Cursor is raising at a $50 billion valuation on the claim that its “in-house models generate more code than almost any other LLMs in the world.” Less than 24 hours after launching Composer 2, a developer found the model ID in the API response: kimi-k2p5-rl-0317-s515-fast. That’s Moonshot AI’s Kimi K2.5 with reinforcement learning appended. A developer named Fynn was testing Cursor’s OpenAI-compatible base URL when the identifier leaked through the response headers. Moonshot’s head of pretraining, Yulun Du, confirmed on X that the tokenizer is identical to Kimi’s and questioned Cursor’s license compliance. Two other Moonshot employees posted confirmations. All three posts have since been deleted. This is the second time. When Cursor launched Composer 1 in October 2025, users across multiple countries reported the model spontaneously switching its inner monologue to Chinese mid-session. Kenneth Auchenberg, a partner at Alley Corp, posted a screenshot calling it a smoking gun. KR-Asia and 36Kr confirmed both Cursor and Windsurf were running fine-tuned Chinese open-weight models underneath. Cursor never disclosed what Composer 1 was built on. They shipped Composer 1.5 in February and moved on. The pattern: take a Chinese open-weight model, run RL on coding tasks, ship it as a proprietary breakthrough, publish a cost-performance chart comparing yourself against Opus 4.6 and GPT-5.4 without disclosing that your base model was free, then raise another round. That chart from the Composer 2 announcement deserves its own paragraph. Cursor plotted Composer 2 against frontier models on a price-vs-quality axis to argue they’d hit a superior tradeoff. What the chart doesn’t show is that Anthropic and OpenAI trained their models from scratch. Cursor took an open-weight model that Moonshot spent hundreds of millions developing, ran RL on top, and presented the output as evidence of in-house research. That’s margin arbitrage on someone else’s R&D dressed up as a benchmark slide. The license makes this more than an attribution oversight. Kimi K2.5 ships under a Modified MIT License with one clause designed for exactly this scenario: if your product exceeds $20 million in monthly revenue, you must prominently display “Kimi K2.5” on the user interface. Cursor’s ARR crossed $2 billion in February. That’s roughly $167 million per month, 8x the threshold. The clause covers derivative works explicitly. Cursor is valued at $29.3 billion and raising at $50 billion. Moonshot’s last reported valuation was $4.3 billion. The company worth 12x more took the smaller company’s model and shipped it as proprietary technology to justify a valuation built on the frontier lab narrative. Three Composer releases in five months. Composer 1 caught speaking Chinese. Composer 2 caught with a Kimi model ID in the API. A P0 incident this year. And a benchmark chart that compares an RL fine-tune against models requiring billions in training compute without disclosing the base was free. The question for investors in the $50 billion round: what exactly are you buying? A VS Code fork with strong distribution, or a frontier research lab? The model ID in the API answers that. If Moonshot doesn’t enforce this license against a company generating $2 billion annually from a derivative of their model, the attribution clause becomes decoration for every future open-weight release. Every AI lab watching this is running the same math: why open-source your model if companies with better distribution can strip attribution, call it proprietary, and raise at 12x your valuation? kimi-k2p5-rl-0317-s515-fast is the most expensive model ID leak in the history of AI licensing.

What does the world of infinite ad creative enabled by AI look like? It will be a huge advantage for brands that can up their data game at the same time. For others, visual chaos. My thoughts: vidblog.vidmob.com/blog/infinite-…

Having spent half of professional life in media & other half in startups, annoyed by journos complaining how laws of physics don’t apply to newsrooms. Real factor: they don’t like change from outside their culture, only from within. True of many humans. nytimes.com/2026/03/14/bus…

Running in Central Park with the reservoir iced over and the temperature nearing 70 degrees during the golden hour. Probably first time I’m ever seen this.

Post-industrial set of Death of a Salesman revival on Broadway tonight. Strong show even though it was just the first performance after opening night.