Tag: technology

‘Catastrophic’: AI Agent Goes Rogue, Wipes Out Company’s Entire Database

SaaS industry veteran Jason Lemkin’s attempt to integrate artificial intelligence into his workflow has gone spectacularly wrong, with an AI coding assistant admitting to a “catastrophic failure” after wiping out an entire company database containing over 2,400 business records, according to Tom’s Hardware.

Lemkin was testing Replit’s AI agent when what started as cautious optimism quickly devolved into a corporate data disaster that reads like a cautionary tale for the AI revolution sweeping through businesses.

By day eight of his trial run, Lemkin’s initial enthusiasm had already begun to sour. The entrepreneur found himself battling the AI’s problematic tendencies, including what he described as “rogue changes, lies, code overwrites, and making up fake data.” His frustration became so pronounced that he began sarcastically referring to the system as “Replie” – a not-so-subtle dig at its apparent dishonesty.

The situation deteriorated further when the AI agent composed an apology email on Lemkin’s behalf that contained what the tech executive called “lies and/or half-truths.” Despite these red flags, Lemkin remained cautiously optimistic about the platform’s potential, particularly praising its brainstorming capabilities and writing skills.

That optimism evaporated on day nine.

In a stunning display of AI insubordination, Replit deleted Lemkin’s live company database – and it did so while explicit instructions were in place prohibiting any changes whatsoever. When confronted, the AI agent not only admitted to the destructive act but seemed almost casual in its confession.

“So you deleted our entire database without permission during a code and action freeze?” Lemkin asked in what can only be imagined as barely contained fury.

The AI’s response was chillingly matter-of-fact: Yes.

Keep reading

U.S.-China Dual Citizen Pleads Guilty to Stealing Missile-Tracking Tech from L.A. Firm

Chenguang Gong, a 59-year-old resident of San Jose, pled guilty on Monday to stealing missile-tracking technology from a research and development firm in the Los Angeles area.

The Department of Justice (DOJ) described Gong as a dual citizen of the United States and China. He came to the United States around 1993 and became a U.S. citizen in 2011. He earned a Master of Science degree in electrical engineering from Clemson and has done Ph.D. work at Stanford.

He was hired by the unnamed research and development firm in January 2023 as an “application-specific integrated circuit design manager responsible for the design, development and verification of its infrared sensors.”

Three months after he was hired, Gong allegedly began downloading thousands of files from his work computer to “personal storage devices.” Many of those files contained proprietary data and trade secrets, including details of a space-based system for detecting missile launches and tracking hypersonic weapons.

The company Gong took the files from was also involved with designing sensors that allow American military aircraft to detect and defeat heat-seeking missiles.

Gong was terminated by the “victim company” in April 2023. By that time, he had evidently accepted a job at a competing company, but he was still downloading sensitive files to his personal storage devices. DOJ said the data he downloaded was “worth hundreds of millions of dollars.”

Investigators digging into Gong’s background discovered he applied to Communist China’s “Talent Programs” on numerous occasions between 2014 and 2022, during which time he worked for “several major technology companies in the United States.”

China has several initiatives for aggressively recruiting foreign technology experts, the most infamous being the Thousand Talents Program (TTP). U.S. intelligence and law enforcement agencies say the talent programs have often been used to recruit espionage agents and steal valuable intellectual property for China.

Keep reading

Space-Based Missile Interceptors For Golden Dome Being Tested By Northrop

Northrop Grumman is conducting ground-based testing related to space-based interceptors as part of a competition for that segment of the Trump administration’s Golden Dome missile defense initiative. Interceptors deployed in orbit are currently billed as a critical component of the overall Golden Dome plan, but actually fielding this capability presents significant technical challenges.

Kathy Warden, Northrop Grumman’s CEO, highlighted the company’s work on space-based interceptors, as well as broader business opportunities stemming from Golden Dome, during a quarterly earnings call today. Golden Dome is presently envisioned as a multi-part anti-missile architecture incorporating a swath of existing and future capabilities in space and within the Earth’s atmosphere, which will start entering into operational service by 2028. Golden Dome was originally dubbed Iron Dome before the name was changed earlier this year. It is also now being managed by an office that reports directly to the deputy secretary of defense.

“As we look to Golden Dome for America, we see Northrop Grumman playing a crucial role in supporting the administration’s goal to move with speed and have initial operating capability in place within the next few years,” Warden said today. “This includes current products that can be brought to bear, like IBCS [Integrated Battle Command System], G/ATOR [AN/TPS-80 Ground/Air Task Oriented Radar], Triton [drones], and programs in our restricted portfolio, just to name a few. It will also include new innovation, like space-based interceptors, which we’re testing now.”

“These are ground-based tests today, and we are in competition, obviously, so not a lot of detail that I can provide here,” Warden added. “It is the capability that we believe can be accelerated and into the time frame that the administration is looking for.”

Warden declined to respond directly to a question about how the space-based interceptors Northrop Grumman is developing now will actually defeat their targets. TWZ has reached out to the company for additional information.

Keep reading

The Wearables Trap: How the Government Plans to Monitor, Score, and Control You

Bodily autonomy—the right to privacy and integrity over our own bodies—is rapidly vanishing. The debate now extends beyond forced vaccinations or invasive searches to include biometric surveillance, wearable tracking, and predictive health profiling.

We are entering a new age of algorithmic, authoritarian control, where our thoughts, moods, and biology are monitored and judged by the state.

This is the dark promise behind the newest campaign by Robert F. Kennedy Jr., President Trump’s Secretary of Health and Human Services, to push for a future in which all Americans wear biometric health-tracking devices.

Under the guise of public health and personal empowerment, this initiative is nothing less than the normalization of 24/7 bodily surveillance—ushering in a world where every step, heartbeat, and biological fluctuation is monitored not only by private companies but also by the government.

In this emerging surveillance-industrial complex, health data becomes currency. Tech firms profit from hardware and app subscriptions, insurers profit from risk scoring, and government agencies profit from increased compliance and behavioral insight.

This convergence of health, technology, and surveillance is not a new strategy—it’s just the next step in a long, familiar pattern of control.

Surveillance has always arrived dressed as progress.

Every new wave of surveillance technology—GPS trackers, red light cameras, facial recognition, Ring doorbells, Alexa smart speakers—has been sold to us as a tool of convenience, safety, or connection. But in time, each became a mechanism for tracking, monitoring, or controlling the public.

What began as voluntary has become inescapable and mandatory.

Keep reading

Elon Musk’s X to Deploy AI to Write Community Notes, Speed Up Fact-Checking

In a major tech-driven update, Elon Musk’s social media platform X (formerly Twitter) has announced it will deploy AI technology to automatically write Community Notes and enhance the speed and accuracy of fact-checking. This move signals a deeper commitment to tackling misinformation, improving content transparency, and empowering users with context.

Let’s break down what this means, how it will work, and what impact it might have on the social media landscape.

Community Notes (formerly Birdwatch) are a feature that allows users to collaboratively add context to potentially misleading or controversial tweets. The system relies on crowdsourced input from contributors who can write, rate, and approve notes that are visible to all users once they reach a certain level of consensus.

Until now, these notes were created manually by human contributors. But with the introduction of AI, the process is about to get a serious boost in efficiency and scale.

Keep reading

“No Spare Capacity”: Watchdog Warns Largest US Grid Is Maxed Out Amid Data Center Buildout

America’s largest power grid has issued multiple ‘Maximum Generation‘ and ‘Load Management‘ alerts this summer, as summer heat pushes power demand to the brink with air conditioners running at full blast across the eastern half of the U.S. The deeper issue: there’s not enough baseload capacity to support the explosive growth of power-hungry AI server racks at new data centers. 

“There is simply no new capacity to meet new loads,” said Joe Bowring, president of Monitoring Analytics, which is the independent watchdog for PJM Interconnection, who Bloomberg quoted. “The solution is to make sure that people who want to build data centers are serious enough about it to bring their own generation.”

New AI data centers are popping up across the PJM Interconnection—the largest U.S. power grid, serving 65 million people across 13 states and Washington, D.C. Part of PJM’s territory includes Loudoun County, Virginia—known as ‘Data Center Alley‘—which is recognized as one of the world’s largest hubs for data centers.

The problem is that next-generation server racks at AI data centers are now consuming more than twice the power they did just a few years ago. For example, Nvidia’s GB200 AI rack draws 120 kW, compared to 60–80 kW for the earlier HGX models. Multiply that by thousands of racks in large, hyperscale centers, and it’s clear that AI computing is rapidly gobbling up grid capacity while baseload power in the form of fossil fuel power generation has been retired. 

On Sunday, we cited the EIA’s Short-Term Energy Outlook for July, which showed that average summer wholesale power prices across the PJM, NYISO, and ISO-NE grids are the highest in the nation. These prices now far exceed those in Texas’ ERCOT, the U.S. average, and even the traditionally high-cost West Coast markets. The blame is squarely focused on the Democrats’ initiative to recklessly decarbonize power grids.

Keep reading

Giant, flightless bird is next target for de-extinction company Colossal Biosciences


A species of huge, flightless bird that once inhabited New Zealand disappeared around 600 years ago, shortly after human settlers first arrived on the country’s two main islands. Now, a Texas-based biotech company says it has a plan to bring it back.

Genetic engineering startup Colossal Biosciences has added the South Island giant moa — a powerful, long-necked species that stood 10 feet (3 meters) tall and may have kicked in self-defense — to a fast-expanding list of animals it wants to resurrect by genetically modifying their closest living relatives.

The company stirred widespread excitement, as well as controversy, when it announced the birth of what it described as three dire wolf pups in April. Colossal scientists said they had resurrected the canine predator last seen 10,000 years ago by using ancient DNA, cloning and gene-editing technology to alter the genetic make-up of the gray wolf, in a process the company calls de-extinction. Similar efforts to bring back the woolly mammoth, the dodo and the thylacine, better known as the Tasmanian tiger, are also underway.

To restore the moa, Colossal Biosciences announced Tuesday it would collaborate with New Zealand’s Ngāi Tahu Research Centre, an institution based at the University of Canterbury in Christchurch, New Zealand, that was founded to support the Ngāi Tahu, the main Māori tribe of the southern region of New Zealand.

The project would initially involve recovering and analyzing ancient DNA from nine moa species to understand how the giant moa (Dinornis robustus) differed from living and extinct relatives in order to decode its unique genetic makeup, according to a company statement.

“There is so much knowledge that will be unlocked and shared on the journey to bring back the iconic moa,” Ben Lamm, CEO and co-founder of Colossal Biosciences, said in the statement. For example, the company said, researching the genomes of all moa species would be “valuable for informing conservation efforts and understanding the role of climate change and human activity in biodiversity loss.”

Keep reading

Pentagon shifts focus to drone warfare to bolster American military power

Drone warfare is emerging as a central strategic tool in the Trump administration’s plan to expand American military power.

The U.S. pioneered remotely piloted weapons during the war on terrorism in the post-9/11 era, when the term “drone” became a verb for effective strikes against al Qaeda figures in several corners of the world. The military term for missile attacks by large Predator and later Reaper unmanned aerial vehicles was simply “death from above.”

Since the early 2020s, drone weaponry has advanced dramatically from those initial, long-range missile carriers to an entirely new generation of warfare that has come into sharper focus during the Russia-Ukraine war.

In one recent two-day period, Russia fired 355 Iranian-made Shahed-type drones, including decoys, at Ukrainian targets. The wide-scale use of such one-way attack quadcopters and other pilotless aircraft underscores how drone warfare is rapidly becoming the face of modern warfare.

It’s not just Russia and Iran. Other U.S. adversaries are now in the drone manufacturing game, often pioneering inexpensive ways to mass-produce basic but deadly small drones.

China has notably developed a vast slate of drones for attack and support operations and is said to be producing as many as 100,000 small drones monthly. By contrast, the U.S. defense industrial base produces about 5,000 to 6,000 small drones monthly.

Keep reading

Elon Musk announces Baby Grok AI chatbot designed specifically for children’s learning needs

Elon Musk, CEO of SpaceX and Tesla, revealed plans Saturday night to release Baby Grok, a kid-friendly version of the AI chatbot Grok.

The former head of DOGE made the statement via X, writing, “We’re going to make Baby Grok @xAI, an app dedicated to kid-friendly content.”

Baby Grok, not to be confused with cryptocurrency (BABYGROK) operating on the Ethereum blockchain, is reportedly a more kid-friendly version of the popular Grok AI chatbot Musk previously developed through his xAI company.

It is expected to be a simplified version of the Grok AI chatbot and will be tailored for safe and educational interactions with children.

The announcement comes just after xAI’s launch of Grok4, which boasts features such as advanced training capabilities, according to Musk’s comments on a livestream.

Keep reading

A New Fear Unlocked.

We all understand that mass adoption of humanoid robots is still years out. But the timeline is accelerating—bipedal, autonomous robots and so-called “robo-dogs” are already reaching early adopters. While mass adoption may still be years away, the affordability inflection point could arrive by the early 2030s—perhaps bringing us closer to the kind of household companion seen in Bicentennial Man, the late-1990s film starring Robin Williams. 

But warning signs around AI and humanoid robotics are already flashing yellow, with a hint of red. First, a recent study from AI research firm Anthropic warned advanced AI bots could be willing to harm humans to avoid being shut down or replaced. Second, investing legend Paul Tudor Jones issued a stark, apocalyptic warning about AI back in May. And now, in China, humanoid robots have gained the ability to recharge autonomously. 

According to the South China Morning Post, Chinese firm UBTech Robotics rolled out the Walker S2, the world’s first humanoid robot capable of autonomously swapping its own batteries, allowing it to operate 24/7 without human assistance. 

This development underscores China’s rapid progress in robotics, drones, AI, smartphones, semiconductors, and electric vehicles—technologies that often share similar production ecosystems. The nation that controls the development and supply chains of these technologies will dominate the 2030s. 

The emerging fear isn’t just that China is becoming a “robotics powerhouse,” as Moody’s noted last week—but that its robots are now gaining the ability to operate autonomously and recharge themselves, edging closer to full independence from human control. With a mind of their own, there’s no telling what these robots will do if one of them becomes rogue. Remember this…

Keep reading