 Your new post is loading...
 Your new post is loading...
Lehi officially completed its city-owned, open-access fiber optic network in August 2025, marking a major milestone for the community after a year of work. The extensive infrastructure project, built and operated in a public-private partnership with Strata Networks, concluded its main physical rollout over the winter, stringing more than 400 miles of fiber and spanning up to 2.1 million feet of cable. The network successfully brings next-generation connectivity and multi-gigabit speeds up to 10-Gig to approximately 23,000 homes and businesses across the city. While engineering and construction milestones have been celebrated as a technical success, local attention has shifted toward consumer adoption rates and the network’s long-term financial viability.
The Capytistas, a secretive group claiming more than $1.2 million in backing, started with Elon Musk and Tesla’s board. Members argue that those who buy political power deserve the scrutiny they’ll impose. An underground activist group says it plans to start openly doxing billionaires as part of a phased campaign against what it calls the American oligarchy—beginning this month with a release that maps the homes of Elon Musk, his brother Kimbal, and Tesla’s board, alongside investors and suppliers that keep the corporation afloat. Two of its members, interviewed by WIRED on the condition of anonymity, say they spent the past year constructing dossiers on billionaires—and one trillionaire—and the people they consider their top lieutenants, using nearly every method available to a private investigator, and some that are not. The pair say the group is international and includes grassroots activists, researchers, journalists, and artists, including some who received top honors in print, television, and digital media in their former careers.
-
OpenAI wants to rebrand AI data centers using a new, more alliterative name. -
"Compute campuses" is their new-fangled term. -
An OpenAI spokesperson said it better represents the "role" of AI infrastructure. OpenAI doesn't want to call them "data centers" anymore. Instead, the ChatGPT maker plans to propose a new, more chic name for the behemoth factories that power your chatbot queries. That new name is "compute campuses," an OpenAI spokesperson confirmed to Business Insider on Sunday.
The average American household paid a record $217 for electricity this July. That was no anomaly: Residential electricity prices climbed 37% in the last decade. Yet over the same decade, wholesale power prices barely moved in most places across the US. Take Oklahoma, where wholesale electricity cost $25 per megawatt-hour in 2015 and remained $25 in 2025. Compare that to natural gas prices, which rose from around $2.50 to $3.50 over the same period, an increase of 40%. So what’s behind the rising bills? Lee Taylor, co-founder and CEO of Resurety and a recent Energy Empire guest, points to a long list of causes: disruptive weather events, labor costs, resiliency efforts, and capacity costs driven up by AI. Understanding them starts with understanding what makes up your bill.
While more cities are canceling their contracts with Flock Safety, just as many are discovering how difficult that can be in some instances. Plymouth, Minnesota’s city council voted to terminate its contract with Flock in September, but failed to respond to the city’s request. For over a month, the cameras remained up, and the city had received no acknowledgment from the company. So, Police Chief Eric Fadden took matters into his own hands and physically disconnected the cameras. “After hearing nothing for more than a week, I directed our staff to disconnect the power cords from the cameras in order to be sure that they are not functioning or recording plates,” Fadden told The Drive.
After years of staged demos, circular financing, and promises of god-like intelligence, the AI industry is facing a brutal reality check. From consumers to governments, the bill is coming due—and the technology's actual utility is far smaller than the trillion-dollar narrative claimed. he "magic-and-miracles" era of artificial intelligence is over. What remains is a market correction of historic proportions—one that is exposing how a coalition of tech executives, venture capitalists, and compliant politicians convinced the world to pour trillions of dollars into a technology whose strongest use cases are drafting emails, summarizing documents, and generating code. The shift from speculative excitement to rigorous scrutiny has been swift. The conversational narrative around AI is no longer dominated by sweeping promises of imminent Artificial General Intelligence (AGI). Instead, it focuses on mounting corporate debt, dismal returns on investment, and the logistical headaches of real-world integration. The "AI bubble" is deflating, separating overhyped software wrappers from genuinely sustainable business tools. What follows is a reckoning across every constituency the industry courted—consumers, businesses, governments, its own workforce, and the investors who bankrolled the entire enterprise. Each was sold a different story. Each is now discovering the far more modest reality.
Previously, Did AT&T Sucker-punch Reporters, the Public, Investors, Politicians, and Regulators? 1% Copper Lines Left? Statements Made by AT&T and Verizon to Investors, 2012–2016: AT&T: 2013 “In the 25 percent of AT&T’s wireline customer locations… it will utilize its expanding 4G LTE wireless network — as it becomes available — to offer voice and high-speed IP Internet services.” Verizon: 2012 “And in many areas, we’re also taking customers that aren’t performing well on copper and we’re moving them over to the wireless technology. So that improves our cost structure significantly and streamlines all those ongoing maintenance costs.” Verizon: 2016 “Once you go wireless, you don’t have to run co-ax, you don’t have to do any of those high labor-intensive activities and so you light up service overnight.” ================================= That’s Right. By 2012–2013, AT&T and Verizon would start to divert billions of dollars of the state utility construction budgets to pay for wireless buildouts, leaving the rural areas to deteriorate. But, there’s a rub because both of these holding companies had state obligations to maintain and upgrade the entire wireline networks because they were and still are utilities. And in every state there were changes in state laws to charge the local phone and all wireline customers extr a fornetwork upgrades — which they never received. And here. in the 2012–2013 time-frame, they told investors about these plans, and this has now come full circle.
Long before he was building broadband across rural Pennsylvania, Dave Gibbons was disrupting monopolies in a college dorm room. As a freshman at Gettysburg College in the early 2000s, he and a fellow student launched an online marketplace for students to sell their old textbooks directly, bypassing the campus bookstore, which bought low and sold high. When the store tried to block them from advertising, he pushed back and won. “Sticking it to the man is definitely part of the ethos,” he laughs. Two decades later, that same instinct to tackle practical problems and challenge incumbents has delivered a significant broadband win in rural central Pennsylvania. Black Bear Fiber founder Dave Gibbons on how the Appalachia Digital Accelerator helped secure nearly $30M in USDA and BEAD funding for rural fiber.
The trillion dollar Anthropic IPO will enrich Google and Amazon, who are ostensibly competitors. How does that make sense? It doesn’t, unless Anthropic is four flailing monopolies in a trench coat. The past month has seen some signs in the financial markets that the AI bubble is beginning to deflate. First it was the neoclouds, which are off-balance sheet entities for large tech and chip companies, getting crushed. Then on Monday, AI giant Anthropic made public its initial draft S-1 filing, the form it has to submit before it sells shares on the stock market. Then, a few days later, Google and Amazon released their earnings, disappointing investors with misses in revenue targets and higher than expected negative free cash flow. Google stock alone fell by 12% on Thursday, a half a trillion dollar collapse. The next day, the whole tech sector got crushed, and it now down 10% from its high. Wall Street, it seems, is worried.
Two cryptocurrency tracking firms uncover evidence that a North Korean state-sponsored group, TraderTraitor, is behind the $6.3 billion hack of Binance, a crypto exchange. Two blockchain-tracking firms, Chainalysis and TRM Labs, point the finger at the North Korean state-sponsored group TraderTraitor, also known as Lazarus, which has targeted cryptocurrencies. TraderTraitor was blamed for last year’s $1.5 billion hack on the Bybit exchange, then the record holder for the largest crypto heist. Last month, the group was also blamed for another $387 million heist on the Bitget exchange.
The League of Minnesota Cities has a FAQ for communities thinking about data center issues... The demand for new data centers is sweeping the nation. At first, data centers may seem like a typical land use application. However, as city leaders consider proposals for their communities, they may soon find themselves faced with complex questions related to environmental regulations, water usage, and onsite or off-site electric generation. Get answers to FAQs regarding data centers
RICHLAND COUNTY — A multiyear effort to bring high-speed fiber internet to underserved rural residents in northern Richland County has officially crossed the finish line, delivering connections to hundreds more addresses than originally anticipated. Ashley Karlen, director of government affairs for Charter Communications, told the Richland County Board of Commissioners Thursday the buildout was completed Oct. 6. The initiative brought high-speed fiber broadband to 1,167 serviceable locations — surpassing the 748 locations initially contracted under the agreement. Charter’s overall service footprint for the project encompasses 1,211 passed addresses, according to follow-up data provided by Karlen.
A $6.3 billion heist exposes crypto's dependence on financial gatekeepers, the limits of Binance's safety net, and a potential payday for North Korea. Thirty Minutes Shortly after four Thursday afternoon, Eastern time, five wallets controlled by Binance, the world’s largest cryptocurrency exchange, began hemorrhaging digital assets. The first transfers totaled roughly $300 million. Then the attackers struck again and again across ten blockchains. Within half an hour, nearly all of $6.3 billion was gone. A smaller wave followed hours later. To put that in perspective, the previous record for a crypto theft was the $1.5 billion North Korean hackers took from the exchange Bybit in February 2025. This one is more than four times larger. Elliptic, a blockchain analytics firm, calls it almost certainly the largest theft of any kind in history, surpassing even the roughly $1 billion Saddam Hussein’s regime hauled out of Iraq’s central bank in 2003. Saddam needed trucks. These thieves needed access to Binance’s transaction systems and half an hour. Binance quickly froze withdrawals and assured its 335 million registered users that their money was safe. That reassurance is the real story here. Cryptocurrency was sold as a way to escape banks and financial gatekeepers. Binance became one of the largest gatekeepers of all. So when that money vanished, who actually stood behind the promise to get it back? The answer says a great deal about what crypto has become.
|
"There are no exceptions that would allow this major source of air pollution to evade the law, and citizens have the right to defend themselves in Court against illegal pollution." GREENVILLE, Miss. — On Friday, the U.S. District Court for the Northern District of Mississippi denied xAI’s motion to dismiss a Clean Air Act lawsuit brought against the company by the NAACP. The lawsuit seeks to hold the tech company accountable for its continued operation of an illegal, unpermitted power plant that supports its Colossus 2 data center. The ruling comes shortly after the Court denied a similar motion from the U.S. Department of Justice.
With the rapid growth in data centers, federal policymakers have been focused on how to make sure that adding large data centers to the electric system does not increase electricity bills for other customers. Recently, a bipartisan group of senators introduced an energy permitting reform bill, the Bipartisan American Affordability and Jobs Act (BAAJA), which would be a major shift in the current paradigm for transmission pricing policies. BAAJA would require electric utilities to charge data centers more for transmission service and use those extra costs to reduce transmission charges to other customers. BAAJA also offers additional policies through its “Ratepayer Protection Floor” that could be implemented by states to ensure further ratepayer protection. Grid Strategies estimates $55 billion in transmission costs could be collected from data centers under BAAJA.
You may have noticed the headline earlier this year in May that said that bot traffic had surpassed human traffic on the Internet. The bigger news since then is that, in a matter of months, bot traffic has grown to be nearly two-thirds of the traffic on our networks. Mitch Wagner wrote an opinion piece in Fierce Network that describes the foreseeable future impact of the growth of machine-to-machine traffic on our networks. It’s not a future that is going to be comfortable for the many small ISPs I work with. Wagner correctly observes that the growth of bot traffic poses a problem for telcos that have built their networks to provide services to people and companies. I remember, early in my career, conducting traffic studies to understand the busy-hour on clients’ voice networks to help them make sure they had sufficient capacity. That’s still standard practice in sizing backbone connections and in designing fiber network electronics. However, the growth of bot traffic changes the equation of sizing and managing networks.
Artificial intelligence chatbots are increasingly becoming a regular part of life for American youth. Two new surveys show that nearly two-thirds of tweens and teens use AI chatbots. Most kids say they use chatbots for help with homework and when they're stressed, lonely, or anxious. And how they use chatbots is tied to their mental health. One survey, done by the online safety advocacy group, Common Sense Media, found that AI use starts early – years before kids hit the minimum age of 13 on most popular chatbots. Nearly half 9-to-12-year-olds have used an AI chatbot or companion. By the time kids are 13 years old, nearly three-quarters of them have used chatbots.
Communities are weighing the costs of new data centers against promises of jobs. An MIT study looks at the employment record behind those claims. As more and more data centers pop up to accommodate increased AI use, there have been a lot of questions regarding who benefits from them. Though the facilities themselves don't employ many people once open, proponents claim that data centers can boost the local economy by creating lots of new jobs. A study from the MIT Center for Real Estate provides some useful insight, revealing that the truth is, unsurprisingly, more complicated. While it does link data centers to "a gradual increase in local employment following data center entry," the research shows that modest success doesn't transfer across the board.
I spoke with Ali Golshan, a senior engineer behind Nvidia's new Open Agent Safety Platform, which promises to revolutionize agentic AI safety. | Edition #340 Making AI Agents Safe As I have discussed in this newsletter, agentic safety is among the most urgent challenges in AI, especially in light of recent events such as the OpenAI/Hugging Face incident, the leak of PHP vulnerabilities, and the rise of AI-powered hacks. A few days ago, I was invited by Nvidia to attend its GTC conference in Berlin, where I had the chance to chat with some of the senior engineers leading its new AI safety architectures and platforms. One of the most anticipated AI safety releases in recent weeks was Nvidia’s Open Agent Safety Platform, which includes application, runtime, and infrastructure layers to ensure that policy can be verifiable and that AI agents can be continuously contained and monitored. To put this in the context of the broader AI safety debate, a few days ago I spoke about the agentic AI safety crisis and the three avenues we must address simultaneously to solve this crisis. One of the avenues I mentioned was the technological one.
This is the first time a company has come to market warning of ‘catastrophic or existential risks to humanity’. Anthropic, developer of AI chatbot Claude, has filed its prospectus and is expected to go public on Wall Street in November at a valuation of up to US$2 trillion. That would make it the highest-valued initial public offering (IPO) ever, ahead of SpaceX’s US$1.77 trillion listing in June. OpenAI, the maker of ChatGPT, is expected to follow next year. While these AI-powered IPOs are massive, they reflect an ever bigger trend: holding AI risk is becoming unavoidable for most investors. How much of your portfolio depends on how well AI does? If you invest in an S&P 500 index fund, the answer is about 44%. In addition, most of the market’s gains right now come from AI. Just six companies – Apple, Microsoft, Meta and chipmakers Nvidia, Micron and AMD – delivered 57% of the S&P 500’s gains since April. Without them, the index would have risen about 4% rather than 10%. For investors or even future retirees, that AI exposure is already in their portfolios and pension funds.
Musk has the rockets, the satellites and the money. But without massive amount of spectrum, he has little chance of challenging the Big Three on the ground. Elon Musk is spending billions buying American radio spectrum. How much more could he acquire, what would it take to challenge the Big Three, and could the economics of building a fourth mobile network become his biggest problem?
The prototype satellite for Project Suncatcher was able to beam back a short haiku about being in space, generated by Google’s Gemma 4 open-source AI model. Google just hit a major milestone in its “moonshot” project to deploy functional, solar-powered data centers in space. The tech giant has now received its first AI response from one of its Tensor Processing Unit (TPU) chips launched into low Earth orbit. During a ground-station pass, Google researchers prompted the company’s Gemma 4 open-source AI model to generate a haiku about being in space. As the satellite passed by Svalbard, Norway's Arctic archipelago, it was able to beam back a short poem: “Blue marble below Data streams in silent grace Starlight feeds my core.”
OpenAI's CEO isn't a fan of the growing trend of running AI models locally on PC hardware, calling it a waste of precious DRAM. 'You want maximum utilization,' he says. It’s not exactly a surprise, but OpenAI CEO Sam Altman isn’t a fan of people running AI models locally because it can pull precious memory supplies away from data centers. “You want maximum utilization in the cloud if you’re trying to deal with this very supply-constrained environment. And somehow, many people’s intuition is the opposite of that,” Altman said in a new interview on the Cheeky Pint podcast. In contrast, OpenAI partners, including Nvidia and Microsoft, have been touting the benefits of running AI models on new PC hardware with high memory capacity. Local AI offers better privacy, but it also means consumers and enterprises can ditch cloud-based AI without paying, making the business model a direct threat to OpenAI and other top providers.
Google just lost approximately $500 billion in market value in a single day. Its shares collapsed 12.6% on October 29, one of the largest shareholder-value destructions in history. What makes this more shocking is that Google’s business is growing. Third-quarter revenue increased 25% to $127.6 billion, while Google Cloud revenue jumped 103% to $30.7 billion. Apparently, even doubling your cloud business is no longer enough to please Wall Street.
FCC Chairman Brendan Carr is circulating a proposal to preempt California "carrier of last resort" regulations that AT&T says are slowing its copper retirement plans in the state. The proposal arrives amid an ongoing clash between federal deregulation goals and state consumer-protection authority. The proposal follows recent FCC rules designed to speed up the retirement of power-hungry copper networks and to free up dollars for network upgrades and the deployment of IP-based wireless or fiber alternatives. It also tied into AT&T's FCC-approved plan to turn down copper services in certain parts of California that's been getting pushback from state regulators. The proposal aims to ensure that "state obligations cannot be used in a manner that prevents a provider from effectuating a discontinuance that the Commission has approved," the FCC said. Carr believes that the proposed order, if eventually adopted, would preempt state regulations related to copper retirements. However, there would be a comment and reply period that would take place before an FCC vote.
|