I just tried Kimi K2 Thinking and the computer agent. Holy moly! This is another DeepSeek moment!

Tornyol (@tornyolsystems) is building micro-drones that kill mosquitoes. They use smartphone microphones, car park assist sensors, and some clever DSP and control to transform 40-gram toy drones into mosquito killers.

In @nytopinion “A.I. is no less a form of intelligence than digital photography is a form of photography,” the philosopher Barbara Gail Montero writes. “And now A.I. is on its way to doing something even more remarkable: becoming conscious.” nyti.ms/441eE0G

US debt is 370% of GDP. And that doesn’t even include unfunded liabilities in the future.

Longevity is no longer a myth — China is developing pills to extend human life Scientists at Lonvi Biosciences in Shenzhen have unveiled a drug based on grape seed extract, which they claim could extend human life to 100 years or more, The New York Times reports. Until recently, life-extension research in China was dismissed as pseudoscience — but now it’s becoming a state-backed priority and a rapidly growing market. “In the past, only wealthy Americans talked about longevity. Now affluent Chinese are joining the trend,” said Gan Yu, co-founder of the biotech firm Time Pie. Liu Qinghua, Lonvi’s chief technology officer, went even further, stating that “living to 150 years is absolutely achievable.” The topic resurfaced after a leaked conversation between Vladimir Putin and Xi Jinping at the SCO summit, where the two reportedly discussed organ transplantation and radical life-extension technologies.

Nobody scales like China. There will be mind-blowing 600,000 driverless robotaxis in China within five years. That’s AI doing real work, not creating videos. #AutonomousVehicles

A 50-Year Mortgage saves you $300 per month on a $1 million loan. In return, you'd pay 2.5 million in interest instead of 1.2 million. 30-Year @ 6.25% = 6,157/mo 50-Year @ 6.75% = 5,286/mo

Goldman forecasts robotaxi deployment (2.8x) lead by China in 2025 will grow to 18x by 2030 — End 2025: • China 🇨🇳: 5,000 • US 🇺🇸: 1,800 2030: • China 🇨🇳: 632,000 • US 🇺🇸: 35,000

🧮 Google just published DS-STAR A data science agent that can automate a range of tasks — from statistical analysis to visualization and data wrangling. Reads messy files, plans steps, writes and runs code, and verifies itself, reaching state of the art on tough multi file tasks. It lifts accuracy to 45.2% on DABStep, 44.7% on KramaBench, and 38.5% on DA-Code, and holds first place on DABStep as of September-25. Earlier agents lean on clean CSVs and struggle when answers are split across JSON, markdown, and free text. DS-STAR begins by scanning a directory and producing a plain language summary of each file’s structure and contents that becomes shared context. A Planner proposes steps, a Coder writes Python, a Verifier checks sufficiency, a Router fixes mistakes or adds steps, and the loop stops when it passes or reaches 10 rounds. This setup handles heterogeneous data because the summaries surface schema, types, keys, and hints, so plans refer to real fields instead of guessing. On benchmarks the gains are steady, moving from 41.0% to 45.2% on DABStep, 39.8% to 44.7% on KramaBench, and 37.0% to 38.5% on DA-Code. Ablations explain the lift, removing the Data File Analyzer drops hard task accuracy on DABStep to 26.98%, and removing the Router also hurts across easy and hard tasks. Refinement depth matches difficulty, hard tasks average 5.6 rounds, easy tasks average 3.0 rounds, and over 50% of easy tasks finish in 1 round. The framework generalizes across base models, with a GPT-5 version doing better on easy items and a Gemini-2.5-Pro version doing better on hard items. Net effect, DS-STAR reduces the gap between messy data and reliable answers across CSV, XLSX, JSON, markdown, and plain text.

While it is true that the US has achieved substantial economic growth over the last several decades, it is also true that in some important dimensions, economic conditions are now more challenging than they were a generation or two ago. For example, production workers (defined as blue-collar workers, skilled and unskilled) needed to work approximately 7,000-8,000 hours to receive sufficient compensation (which includes the cost of employee-provided dental/health insurance) to purchase the average house in the 1970s-1980s. Today, they need to work 14,000 hours, and the fraction of compensation that goes to health insurance and other non-wage forms of compensation has increased; as a result, the number of hours they need to work to purchase the average house today is about twice that of the 1980s.

It's impossible to understate how much archaic HIPAA rules around PHI slow down the rapid development and prototyping of AI-enabled health-care applications.

Now online in @CD_AACR: Priming with DNMT Inhibitors Potentiates PD-1 Immunotherapy by Triggering Viral Mimicry in Relapsed/Refractory NK/T-cell Lymphoma - by Cheng Huang, Yan Gao, Jianfeng Chen, Choon Kiat Ong, Huiqiang Huang, and Jing Tan doi.org/10.1158/2159-8290.CD…

Mark Zuckerberg and Priscilla Chan are shifting the focus of their philanthropy toward AI-driven biology, via Biohub. Four “grand challenges” are cited: 1. Unified AI-based model of the human cell 2. Next-gen imaging systems to visualize complex biology 3. Instruments to monitor/modulate inflammation in real time 4 . AI-driven immune system reprogramming for disease prevention/treatment biohub.org/blog/frontier-ai-…

Jensen is getting desperate. China is shutting off a major src of demand for Nvidia chips, transshipment to Chinese SOE DCs. I have pretty solid src that just 6000 out of 100k racks in Meta's Hyperion DC will be Nvidia cards. Google TPUs are supplying most of its future demand. Same w/ Amazon & Trainium. I've said this many times now. The absolute compute per card should not be overvalued. You can achieve same w/ compute w/ more lower cost cards & get better performance through more memory + faster interconnect. On top of that, US data build out is facing logjam due to energy & supply chain issues. You can check the lead time on diesel generators & gas turbine. I overstated optical transceiver issues since the big time has hit supply chain constraint b4 we even got there. Jensen knows Nvidia is facing an upcoming cliff. Altman sees the same issue. Hence all the begging for govt help. At end of the day, Chinese AI labs have shown you can do leading models w/o having unlimited compute, so why do we need to keep proclaiming build out speed that's not achievable?

The mandatory veiling of women was once a pillar of the Islamic Republic of Iran, @arashtehran writes. Now it’s almost gone: theatlantic.com/internationa…

this is a zero sum war of U.S...U.S made a mistake...when they let China freely do what they like from platform, AI structure to Chip, HBM....congratulations they lost 60% of their customers.

Autonomous delivery e-bikes: > 25x cheaper than autonomous car > 6x speed and throughput increase over sidewalk robots > favorable vehicle classification for regulatory / insurance > extremely low emissions

Voters are giving the Trump administration credit and blame where it’s due. A majority of voters thinks the Trump admin has done a good job securing the border and combatting unlawful immigration, but an even larger majority thinks they are failing on a host of other issues, most importantly “inflation and the cost of living”! Inflation and the border were the twin issues that brought down the Biden admin. Inflation in fact brought down many incumbent governments worldwide the last couple years. Biden admin tried to browbeat voters that inflation was addressed and to trust govt statistics over their own eyes and lived experiences. That ended very poorly for them. Now the Trump admin seems to be trying the same thing.

Shenzhen alone has so many amazing robotics companies.

The most wonderful travel experience on planet earth. Cruising through the beautiful landscapes of China 🇨🇳 in 350 km per hour. Absolute comfort, eating good food, trains always on time. China the last fifteen years has built 45 000 km high speed rail - that is a distance from Beijing to Paris return, three times.