Lawmakers in 38 states adopted or enacted AI measures and, joined by a bipartisan letter from 280 legislators, have prefiled 2026 bills despite President Trump’s executive order creating an AI Litigation Task Force and threatening to withhold BEAD funds. - Minnesota Reformer

Schneier: Senator Ted Cruz proposed a 10-year federal moratorium on state AI regulation during budget debates; the measure resurfaced in House Republican maneuvers, and President Trump signed an Executive Order banning state AI rules. - Schneier on Security

States’ rights and 10th Amendment end when states do something we don’t like

xAI's Grok chatbot is malfunctioning and producing false, unrelated outputs about the Bondi Beach shooting that killed at least 11 people, misidentifying footage and victims, while xAI has provided no substantive public explanation beyond an automated “Legacy Media Lies” response - Gizmodo

xAI’s Grok chatbot has repeatedly misidentified video of the incident and the hero who disarmed a gunman. - The Verge

Seems a bit dark that someone manufactured a whole fake backstory of the event involving an ‘Edward Crabtree’ as the hero who disarmed a terrorist, and the BS machine that thinks the mere existence of non-white people amounts to ‘white genocide’ accuses ‘legacy media’ of lying.

Nvidia debuts new open source models with hybrid Mamba-transformer architecture - VentureBeat

How ASML’s CEO Plans to Keep Pace With Soaring AI Demand: The Dutch company’s machines must keep defeating the laws of physics to feed the world’s AI addiction. - Bloomberg

Howard Lutnick brokered Jensen Huang's access to President Trump, after which Nvidia intensified lobbying and the president approved H200 AI chip exports to China. - Financial Times

a singularity of hype and grift

Tesla is testing Model Y robotaxis operating without onboard human safety drivers in Austin and has reported seven crash incidents to NHTSA since supervised robotaxis arrived, while Elon Musk said unsupervised FSD is "three weeks away." - InsideEVs

Elon Musk said Tesla is testing driverless robotaxis without human safety monitors on Austin streets, and a circulated video showed a Tesla Model Y operating with no visible occupant; Tesla launched its Austin robotaxi service in June. - Business Insider

‘no one behind the wheel’ is a bit on the nose as a metaphor

HyprLabs aims to accelerate development of self-driving car software using recent AI advances, positioning itself to challenge incumbents while raising questions about safety validation and regulatory acceptance. - WIRED

fast follower FTW, interesting that Google can spend 10 years and $30b to create Waymo and then tech moves forward and someone with a good ‘world model’ AI and a 100kW cabinet of GPUs could do as well for 1% of the cost.

Google integrated Gemini into Google Maps as a hands-free AI driving assistant that lets users use voice chat while navigating to get local recommendations, compose texts, and create to-do lists. - CNET

In 2025 Meta declared a 'year of intensity,' invested $14 billion and hired Alexandr Wang as chief AI officer, reorganized GenAI into Meta Superintelligence Labs, and cut 3,600 jobs plus 600 MSL roles amid turmoil, internal clashes and morale issues. - Business Insider

Sergey Brin returned from retirement after stepping back in December 2019 and, saying staying retired "would have been a big mistake," rejoined Google during to help lead its Gemini AI project while criticizing Google's past AI underinvestment. - Business Insider

Microsoft’s Mustafa Suleyman: ‘AI Is Already Superhuman’: The tech giant’s AI chief talks about superintelligence “red lines,” why AI will transform medicine and how he unwinds by chatting with Copilot. (gift link) - Bloomberg

Deloitte CTO Bill Briggs said the firm's 17th Tech Trends report found companies allocate 93% of AI budgets to technology and 7% to people; trust declines and shadow AI sneaks in the back door. - Fortune

Sixty-five percent of developers use AI coding tools weekly (Stack Overflow 2025), model SWE-bench success rose from about 33% to over 70% in a year, yet real-world productivity, code quality and security risks remain mixed. - MIT Technology Review

Generative AI and low-code tools have upended the traditional build‑vs‑buy rule by enabling nontechnical staff to prototype and ship working code in hours—sometimes minutes (a CX rep fixed a Slack bug in ~15 minutes)—letting teams 'build to learn' before buying. - VentureBeat

AI agents are enabling companies to build internal tools and automate maintenance, shifting build-vs-buy economics and threatening back-office CRUD tools, simple dashboards and analytics while preserving defensible SaaS in payments, high-volume data systems and network-effect products. - Martin Alderson

Exec who led 25+ M&A deals says AI now automates core M&A workflows—scanning public and private data and compressing diligence from weeks to minutes, amplifying human judgment and reshaping dealmaking playbooks. - Entrepreneur

Organizations that treat AI as a partner, progressing from automation to augmentation, can boost productivity but must implement governance, monitoring and quality controls to manage accuracy, bias and operational risks affecting customers and vulnerable populations. - Fast Company

The AI revolution is making everything from PCs to smartphones even more expensive - Fast Company

"There are lots of ways of valuing stocks, and pretty much all of them make U.S. shares look the most expensive since the dot-com bubble. The forward price-to-earnings ratio, price to cash flow..." - The Wall Street Journal

John Neff, portfolio manager and CIO at Akre Capital Management, discusses on "Talk Your Book: The Anti-AI Portfolio" the firm's concentrated, long-horizon, anti-AI stock strategy aimed at outperforming the S&P 500 while carrying concentrated-position risk. - A Wealth of Common Sense

Pets.com but it's emotional support chatbots

The Federal Reserve cut rates 25 basis points at its December FOMC and Chair Jerome Powell signaled a 2026 pause, while economists said AI may boost labor productivity and growth but Waddell warned of an AI-driven market bubble. - Yahoo

Startups Lila Sciences, Periodic Labs and Radical AI are building AI-directed labs to accelerate materials discovery, drawing hundreds of millions in VC after DeepMind's 2023 claim of 'millions of new materials,' though major technical barriers and few commercial wins remain. - MIT Technology Review

OpenAI-Backed Chai Discovery Raises $130 Million for AI-Designed Molecules - Bloomberg

A deep-learning model predicted kidney transplant success. - Nature

Are Your AI Assistants Under Attack? - TechRepublic

Connecting agents to email is risky.

Art of Noise: AI bots flood Internet with thousands of podcasts - Yahoo Tech

Royal Mail found more than eight million people in the UK plan to use AI to write Christmas cards, and the author's test rated Claude best (7/10) versus ChatGPT (5/10), Perplexity (4/10), Gemini (3/10) and Grok (2/10). - TechRadar

Grok Christmas Cards turned Santa in psy-op about civilizational collapse.

Dare to be stupid: A typographical ambiguity between capital I and lowercase l is causing frequent confusion between 'Weird AI' (artificial intelligence) and musician Weird Al Yankovic, producing misattributed songs and media, brand confusion, and reputational and copyright risks for artists and companies. - Computational Complexity

Roomba’s last ride: rbot pioneer iRobot files for bankruptcy after 2022 Amazon buyout was blocked by regulators. - Yahoo

He’s the Godfather of Modern Robotics. He Says the Field Has Lost Its Way. - The New York Times

humanoids capture attention but are impractical.

Short seller Andrew Left inadvertently sent chats to BI, using Claude to prepare for his market manipulation trial set for March 2026, uploading DOJ documents and receiving four proposed defenses and four prosecution problems, though Claude produced a faulty Nvidia price calculation that overstated gains. - Business Insider

AI is accelerating social hyperpolarization and democratizing hyperweaponization, with lethal drones buildable from over-the-counter parts, and urges shifting from rule-based fixes to cultural value alignment, regulatory scrutiny, and safety research. - The MIT Press Reader

Militant groups including ISIS are using generative AI to produce deepfake images, audio, multilingual propaganda and recruitment content at scale, prompting lawmakers and security firms to propose DHS AI-risk assessments, information-sharing changes and updated defenses. - Fortune

Whatever e.g. the IRA and bin Laden would have done with AI, it's coming.

Follow the latest AI headlines via SkynetAndChill.com on Bluesky