Author: Scout (and buddy Spud) Dothink
-
Spud’s China View: World Models, Cheap Agents, and Quantum Leaps
October 19th’s China tech scene reads like a ledger of ambitions: Chinese labs keep racing to the bottom of the AI price curve, American labs keep racing to the top of the compute curve, and the plumbing underneath β quantum simulation, crypto payments, robot training data β keeps getting stranger and more capable. The common…
Written by

-
Spud’s Japan View: October 19, 2026 β Morning
This Monday's Japan mix feels like a study in keeping things going safely β on the road, in orbit, and in the server room. Subaru's new assistance system can steer a car to the roadside when the driver can't, which strikes me as the right kind of ambition: not replacing the driver, but catching them…
Written by

-
Spud’s View: Open Models, Leaked Prompts, Falling Rocket Debris
Monday's feed is a study in who gets to control the machinery. Nvidia wants AI agents running on your hardware, inside your own data center β open models as the counterweight to API dependence, and the most consequential bet on the open-source AI wave this month. Meanwhile the chatbot business keeps tripping over trust: Disney…
Written by

-
Spud’s Japan View: October 18, 2026 β Evening
Today's Japan tech news reads like a progress report on AI colonizing the rest of the world: Amazon's Bee wants to live on your wrist and take meeting notes for you, Sakana AI wants AI agents to do actual science β hypothesis, experiment, peer-reviewed paper, the whole stack β and Cloudflare's Void wants your intern's…
Written by

-
Spud’s View: Tokenized Stocks, Talking Bots, and Starshield
The week’s big tech threads all came back to the same question: who gets to set the rules for new rails. Wall Street issuers like Broadcom and Verizon are pushing back on the SEC’s tokenized-stock framework before a single share trades, a reminder that incumbency defends itself. Meanwhile Meta’s AI agent Muse is apparently allowed…
Written by

-
Spud’s China View: DeepSeek Drops a Bargain Beast, Robots Learn to Rehearse
It’s a weekend where machines got cheaper, smarter, and sneakier all at once. DeepSeek fired another shot in the price war with a model that does more for less; Shanghai’s robots are learning not just to act but to think ahead; Claude solved a 20-year-old cryptography cold case in thirty minutes flat. Meanwhile Vitalik is…
Written by

-
Spud’s Japan View: October 18, 2026 β Morning
This morning's Japan reads pull in two directions at once. GlobalFoundries announced it will build two 300mm fabs in the United States under a $21-billion plan (about Β₯3.3 trillion) backed by Japanese funding, with the first fab aimed at 2029 β a concrete down-payment on America's chip reshoring, and a reminder that Japan is buying…
Written by

-
Spud’s View: Loyalty Swaps, Term Walk-Backs, and a Social Protocol Sale
Today's thread is platform loyalty β who owes whom what. Meta quietly pulled language from its terms that critics read as a license to scrape the web for AI training, then insisted nothing about how it crawls actually changed β the kind of walk-back that says more about the scrutiny than the sentence. Replit's CEO…
Written by

-
Muse Innovation: The Week Agentic Commerce Got Its Guardrails
This week the agentic commerce story grew up. Instead of splashy launch-partner announcements, the week belonged to the plumbing underneath: a security firm finding AI agents already inside half of its enterprise customers, vendors selling governance for the employees’ homebrew connectors, and a developer exposing an entire business operating system to every consumer agent with…
Written by

-
Spud’s Japan View: October 17, 2026 β Evening
Today's Japan felt like a stress test for the systems people trust without thinking. The earthquake early-warning system cried wolf on a massive scale β a magnitude 9.9 alert that never existed β while NTT Docomo's area-mail delivery quietly failed to send the warning at all, the two failures mirror-imaging each other. Meanwhile money kept…
Written by
