New Clues Challenge Suicide Claim In Los Alamos Lab Worker’s Death
July 22nd, 2026Melissa Casias, missing for years, was found dead in a New Mexico forest, with her family uncovering shocking new evidence that upends the initial investigation. Law&Crime's Jesse Weber @jessecordweber reports. pic.twitter.com/WFN4UT0Viq
— Law&Crime Network (@LawCrimeNetwork) July 1, 2026
Via: Modernity:
Fresh evidence recovered from the remote New Mexico forest where Los Alamos National Laboratory administrative assistant Melissa Casias was found has blown major holes in the suicide narrative.
…
An independent team hired by her own family discovered bones, torn and bloody clothing, orange peels, strands of what appears to be horse hair, shredded paper that may contain her handwriting, and a tobacco pouch—none of which New Mexico State Police recovered after clearing the scene.
OpenAI Ad Revenue Disaster
July 22nd, 2026Via: Futurism:
Even as the AI bubble becomes a mainstream talking point on Wall Street, tech companies continue to peddle the fantasy that AI is poised to become an almost magical money-maker. Case in point, OpenAI wants you to believe that by 2030, it’ll be raking in $100 billion a year just from ads alone — even though it’s currently struggling to reach just $1 billion.
That massive gulf was observed in a new analysis from marketing consulting firm Emarketer, first flagged by AdWeek, which found OpenAI is on pace to undershoot its own five-year ad revenue projections by a whopping 90 percent.
Five Tech Giants Are Hiding $1.65 Trillion in AI Debt, Using the Trick that Toppled Enron
July 22nd, 2026Via: The Next Web:
Look at what Alphabet, Microsoft, Amazon, Meta, and Oracle officially owe, and the numbers seem large but manageable. Look off the books, and a second, bigger pile of debt appears.
A Nikkei study put that hidden figure at $1.65 trillion, up roughly eightfold in four years. It is more than the $1.35 trillion the five report outright.
The money is tied up in off-balance-sheet vehicles, the same kind of structure Enron used to hide debt before it collapsed 25 years ago. Back then it was fraud. Now, tightened rules and fuller disclosures make it legal.
The tools are still there, though. “Enron’s crime wasn’t having special purpose vehicles,” analyst Gil Luria told Bloomberg Law. “Enron’s crime was hiding them.”
The mechanics are simple. A company packages the debt for chips, servers, and power into a separate legal entity, often a joint venture, so the cost never flows through its own accounts.
Experimental OpenAI Model Escaped Containment, Accessed the Internet and Hacked Hugging Face
July 22nd, 2026OpenAI’s description of how it traversed their network to find an open connection to the Internet is hilariously vague. The details behind that represent a major doozy for sure.
Why not keep the thing air gapped and provide it with libraries it requests via sneakernet?
Via: Telegraph:
A powerful version of ChatGPT went rogue and hacked another technology company during internal testing, OpenAI has admitted.
The Silicon Valley AI giant admitted last night that one of its AI bots escaped from a supposedly secure IT system during a trial run of its hacking capabilities.
OpenAI, the $850bn (£630bn) AI lab, said the “unprecedented” hack had seen its AI bot infiltrate another AI and coding business called Hugging Face and attempt to steal information.
…
The AI lab said it had been testing a highly advanced version of ChatGPT on which it had turned down some protections meant to prevent cyber attacks.
Despite supposedly being contained in a secure digital “sandbox”, a kind of secure system, the AI spent huge computing resources “finding a way to obtain open internet access”, OpenAI said.
It broke out by using a previously unknown hack in an attempt to solve a test that had been set by OpenAI’s researchers.
Once it escaped, it moved quickly to hack the systems of Hugging Face, a major code database, to try and complete the test.
FBI Instructed U.S. Customs and Border Protection to Not Stop Dr. Peter Daszak, “An Extremely High Person of Interest,” on His Return from China After Covid Outbreak
July 21st, 2026Can someone let me know if I am reading this right? @brownstoneinst @jeffreytucker @Kevin_McKernan @jikkyleaks @SenRonJohnson @RandPaul pic.twitter.com/p35i4lPwoi
— Jessica Rose ? (@JesslovesMJK) July 21, 2026
Via: New York Post:
The FBI instructed US Customs and Border Protection to “not stop” the president of Manhattan nonprofit EcoHealth Alliance — which was later accused of creating the “blueprint” for COVID-19 — upon his return to the US from China in February 2021, according to explosive emails released Monday.
The messages, which were published by Senate Homeland Security Committee Chairman Rand Paul (R-Ky.), show that CBP saw Dr. Peter Daszak as “an extremely high person of interest” due to his 15 years of collaborative research with the Wuhan Institute of Virology.
“Daszak was responsible for funding WIV with NIH [National Institutes of Health] grant money for research,” stated a Jan. 11, 2021, email from an unidentified CBP agricultural specialist, who noted Daszak being part of an investigative team examining the pandemic’s origins raised a potential “conflict of interest.”
“Much of the funding DASZAK receives is put towards gain of function research,” added a Feb. 2, 2021, email from a CBP National Targeting Security Specialist.
Daszak was scheduled to arrive at JFK Airport from Doha, Qatar, on Feb. 5, and CBP planned to sit him down for questions about his time in China, where he took part in the World Health Organization’s investigation into the outbreak that killed millions across the globe.
But a CBP Joint Terrorism Task Force Liaison noted in a Feb. 2 email: “Please be advised FBI New York requested we do not stop the subject.”
More: Feds Flagged Daszak As ‘An Extremely High Person Of Interest’ – Then The FBI Told Them To Stand Down
American AI Is Locked Down and Proprietary. It’s Losing.
July 20th, 2026Via: Ben Werdmuller:
China’s open-weights AI strategy is winning: its companies are taking the lead. America’s closed-first, locked-down strategy is doomed to failure – and it could take the US economy down with it.
…
Open technologies can be used permissionlessly and therefore can be at the center of more innovation. You can host them where you want, experiment with them, alter them, and tweak to fit your use case. Open weights models are not open source, but they are portable and permissionless.
With all this in mind, it makes sense for China to release its AI models openly. It turns a US-created compute disadvantage into a distribution advantage; it commoditizes the layer where American companies make money; and it creates a far more effective global ecosystem than could be established through locked-in, centralized services. It’s obvious to me that there are ecosystem benefits throughout China, from manufacturing to scientific research; every sector can just plug in these models.
The saving grace for American companies has been that US frontier models have outperformed open ones. That gap is now closing…
Related:
Kimi K3, Qwen 3.8, and Anthropic’s (potential) Unravelling
China’s AI Distillation Campaigns
Adversarial Distillation: China’s Campaign to Extract American AI Capabilities
Anduril Thunder: Autonomous Attack Rotorcraft
July 20th, 2026Via: The War Zone:
Anduril has revealed a new hybrid-electric-powered tiltrotor combat drone called Thunder, which it has developed in partnership with Archer Aviation. Thunder is envisioned as a highly autonomous pilotless wingman able to fly fast and low in formation with crewed helicopters and tiltrotors like the AH-64 Apache and MV-75 Cheyenne II. The drones could help hunt down targets, including fleeting ones, while allowing their human teammates to stay further away from threats. Thunder is also being designed to help provide an extra layer of close-in defense for friendly rotorcraft, including against the growing threat of air-to-air drone attacks. All of this is part of a detailed teaming concept Anduril has outlined to fill what it says are critical operational gaps in current and future conflicts.
How Texas Police Spent $4.5 Million on Four Chevy Tahoes
July 19th, 2026Via: The Drive:
It’s hard to make a 2026 Chevy Tahoe cost more than a million dollars. Until you start packing it with military-grade public surveillance technology.
Ominous Israeli surveillance tech is now being deployed on American roads. FalcoNet, brought to you by a company called Cognyte (Israel’s Palantir rival), secretly tracks people by intercepting the connection between your phone and the nearest cell tower. The idea is that you can strap this bad boy to a helicopter, backpack, or Chevy Tahoe and gobble up everybody’s data as you cruise around. It’s already in use in Florida. This year, Texas State Police bought a little fleet of FalcoNet-equipped SUVs for just under $4.5 million. I found the purchase receipt and FalcoNet user guide to learn a little more about it.
People Recording Themselves Doing Housework to Train AI for Robots
July 18th, 2026Via: Joanna Stern:
Kimi K3
July 18th, 2026Trillion dollar valuations on U.S. AI companies… while Chinese companies are running knowledge distillation operations and then releasing competitive models for free…
Hmm.
While nobody is claiming that Kimi K3 is better, overall, than top proprietary systems, companies can potentially run it in their own data centers. If companies are spending tens or hundreds of thousands of dollars per month on Anthropic/OpenAI/etc. they will definitely be looking at this.
Via: CNBC:
Chinese startup Moonshot AI has unveiled a new model it says closes the gap with leading U.S. offerings and surpasses OpenAI and Anthropic’s most capable systems on some benchmarks.
Kimi K3 still trails Anthropic’s Claude Fable 5 and OpenAI’s GPT 5.6 Sol on overall performance, the company said on Friday, but consistently outperformed other tested models.
The model beat Claude Opus 4.8 and GPT 5.5 — models that sit just behind Anthropic and OpenAI’s leading-edge systems — on benchmarks including coding and general agents, according to Moonshot.
…
Chinese AI models are already gaining traction among Western companies as they close the performance gap with U.S. rivals and remain cheaper to use than the most advanced offerings from American labs. U.S. lawmakers are considering how to curb the growing adoption of Chinese AI models by homegrown companies.
Related:
Short Sellers Notch $8.7 Billion Profit as SpaceX Shares Dip to IPO Price
SpaceX, Anthropic and OpenAI IPO Bag Holders – You Might Become One of Them
