Activating Sst-Chodl neurons—about 1% of inhibitory neurons in the mouse visual cortex—can put mice to sleep and increases slow-wave activity, with stimulation raising delta power across every cortical layer and cutting the time to fall asleep.
Sep 18, 2026 ·Ars Technica
CoreWeave’s Q2 revenue was $2.58 billion (~112% year over year) and it reported a $104 billion revenue backlog as of June 30, giving it multi-year visibility even as Cramer said the stock becomes more speculative if rates rise.
Sep 18, 2026 ·
Yahoo Finance
The National Archives removed an open-source Alibaba Qwen AI search tool from the Federal Register website after it was publicly noticed and questioned by the FBI and lawmakers, and the service had been available for at least a day (noted via a September 15 screenshot and Wednesday removal).
Sep 18, 2026 ·Ars Technica
Joby Aviation says an autonomous aircraft flew 3,100 miles across the United States without any human pilot taking control at any point, performing autonomous taxiing, takeoffs, navigation, and landings while being remotely supervised and with a pilot onboard for compliance.
Sep 18, 2026 ·
TechCrunch
Okta reported fiscal second-quarter 2027 revenue of $805 million (+11% year over year) and raised its fiscal 2027 revenue outlook to $3.216 billion–$3.226 billion (10%–11% growth), with total RPO rising 17% to $4.86 billion and current remaining performance obligations up 14% to $2.59 billion.
Sep 18, 2026 ·Yahoo Finance
Anthropic CEO Dario Amodei proposes “pacing the frontier” by using independent safety evaluators and coordinating AI development among labs in democratic countries.
Sep 18, 2026 ·TechCrunch
California Gov. Gavin Newsom issued an executive order directing the state to develop AI safety rules that could include a routinely verified “kill switch” for frontier models within two months of expert recommendations.
Sep 18, 2026 ·
The Verge
Manus is in talks to raise $500 million at a $4 billion valuation after resuming independent operations following its break from a planned Meta merger blocked by Beijing.
Sep 18, 2026 ·TechCrunch
SAG-AFTRA and WGAE say entertainment AI isn’t as existentially dangerous as doomer warnings, but it still threatens jobs and they’ve bargained “enforceable, meaningful guardrails” and AI protections into major employers’ deals, including requirements to gain permission and compensate actors for AI-generated replicas.
Sep 18, 2026 ·The Verge
J.P. Morgan Private Bank found that 65% of global family offices plan to prioritize AI investments despite concerns about inflated valuations and pricing.
Sep 18, 2026 ·TechCrunch
A TechCrunch Disrupt 2026 Builders Stage session on October 13–15 in San Francisco will debate the trade-offs between open and proprietary AI for startups, led by Nvidia’s Nader Khalil and Sydney Sykes.
Sep 18, 2026 ·TechCrunch
Meta’s Muse AI assistant is now available on Mac and can interact with your files, messages, calendar, notes, and mail inside native apps, with opt-in access and approval prompts before sensitive actions.
Sep 18, 2026 ·TechCrunch
Seagate’s 2026 Data Infrastructure Readiness Report found 38% of organizations say they’re prepared for AI-driven storage demand growth, while 62% are ill-prepared.
Sep 18, 2026 ·
ZDNet
Siri Recap on Apple Watch frames Apple’s approach to AI-augmented ambient audio as keeping personal data, audio recordings, and transcripts from being stored, while relying on privacy “guardrails” rather than obtaining consent from everyone who could be exposed.
Sep 18, 2026 ·
Computerworld
Robinhood reported 28.6 million funded customers and $384 billion in total platform assets at the end of August 2026, and in Q2 said Robinhood Banking topped $3 billion in deposits, its credit card business surpassed $100 million in annualized revenue, and prediction markets traded more than 3.5 billion contracts.
Sep 18, 2026 ·TechCrunch
KDE will mark its 30th anniversary at Akademy in Graz, where a Sunday talk proposes an “AI-native” Plasma that compiles itself per user, device, and moment from a small portable user model into an encrypted, vendor-agnostic personal kernel (“Kadai”).
Sep 18, 2026 ·
The Register
Hacktron AI security researchers used Anthropic’s Claude to chain two vulnerabilities in Discourse’s image-processing pipeline and gain access to multiple OpenAI employee ChatGPT accounts, demonstrating that an unflagged ImageMagick/libheif bug (fixed months earlier but without a CVE) could still be exploitable.
Sep 18, 2026 ·TechCrunch
A Pew Research Center survey across 37 countries finds that in 34 of them, more people expect AI to lead to fewer jobs over the next 20 years than more jobs.
Sep 18, 2026 ·Computerworld
Percona CEO Peter Farkas says no existing database is ideal for agentic AI workloads, and that the “best solution” likely won’t exist until AI requirements mature enough to define what a clean-slate database should be built for.
Sep 18, 2026 ·The Register
Anthropic has established a biology research lab in the San Francisco Bay Area to conduct physical experiments, and its head of life sciences says it is already doing in-house real lab work.
Sep 18, 2026 ·
Engadget
