reSee.it - Related Video Feed

Larry Fink, CEO of BlackRock, is described as saying that building the biggest AI data centers in the United States will require “trillions of dollars” of capital, and that governments cannot build them alone due to lack of resources and growing deficits. The transcript claims these data centers are being built without public approval and without public input. A Utah data center is highlighted as an example: the Stratus Data Center in the empty desert of northwestern Utah near Snowville, close to the Idaho border. The project is said to be pushed by Kevin O’Leary. It is described as being more than twice the size of Manhattan and as potentially needing up to three gigawatts of electricity, compared to the output of multiple nuclear reactors. Environmental groups are said to warn it could raise Utah’s planet-warming pollution by nearly fifty percent, and that its power systems could consume up to 16.6 billion gallons of water per year—enough to fill around 25,000 Olympic swimming pools—despite being in one of the driest states in America. The transcript also uses multiple size comparisons (including San Francisco, Disneyland, Disney World, Paris, suburban house lots, Los Angeles to Central Texas, and football fields) and adds that it could raise daytime temperatures by five degrees and nighttime temperatures by 28 degrees. The project is characterized as an “ecological disaster.” The transcript then shifts to a “very emotionally charged” meeting in Box Elder County. Box Elder County commissioners are said to have moved to approve the Sprouts project after protests outside, a crowded exhibit hall, multiple interruptions, and then shifting to a smaller room and broadcasting to Zoom, which upset people. Commissioners are described as saying the county’s land is not zoned, limiting their ability to stop the project, and that approving it allowed them to obtain concessions from the developer. Finally, the transcript questions what so much data would be for, suggests it could be intended for the largest, most expensive AI surveillance system in human history, and links that idea to a claim that Trump and other billionaires traveled to China weeks earlier for deals or negotiations related to AI surveillance, framing this as a conspiracy idea.

- Indianapolis residents organized to stop Google's proposed $1,000,000,000 AI data center on a 500-acre site, which reportedly would have used 1,000,000 gallons of water per day. Google withdrew its petition to build, preventing a city council vote. Community members described the victory as “we beat Google,” while warning the fight isn’t over and noting tactics used by a secretive tech company in Saint Charles, Missouri. Residents voiced fears about water supply, contamination, and rising electricity costs, with one farmer stressing the risk to livelihoods if water is unavailable. - The victory was celebrated as a win for community power, though participants cautioned that Google could reappear with a new plan in a few months. The broader context included concerns that big tech seeks data centers in communities, potentially impacting water and energy prices, and the possibility of revisiting projects once opposition fades. - An NPR overview on America’s AI industry highlighted concerns about data centers depleting local water supplies for cooling, driving up electricity bills, and worsening climate change if powered by fossil fuels. The IEA warns climate pollution from power plants serving data centers could more than double by 2035. In the Great Lakes region, water utilities, industry, and power plants draw from a shared resource; questions arise about how much more water the lakes can provide for data centers and associated power needs. - Examples cited include Georgia where residents reported drinking-water problems after a nearby data center was built; Arizona cities restricting water deliveries to high-demand facilities. The Data Center Coalition notes efforts to reduce water use through evaporative cooling versus closed-loop systems; a Google data center in Georgia reportedly uses treated wastewater for cooling and returns it to the Chattahoochee River. There is a push toward waterless cooling, with a balancing act described: more electricity to cool means less water, and vice versa. - Rising electricity bills are a major concern as data centers increase power demand. A UCS analysis found that in 2024, homes and businesses in several states faced $4.3 billion in additional costs from transmission projects needed to deliver power to data centers. The dialogue includes questioning why centers aren’t built along coastlines where desalination could be used at the companies’ own expense, arguing inland siting imposes greater resource strain on residents. - Financial concerns extend to tax incentives for data centers. GoodJobsFirst.org reports that at least 10 states lose more than $100,000,000 annually in tax revenue to data centers; Texas revised its cost projection for 2025 from $130,000,000 to $1,000,000,000 within 23 months. The group calls for canceling data center tax exemption programs, capping exemptions, pausing programs, and robust public disclosure. - The narrative concludes with a call to resist placing data centers in established communities, urging organized action and advocating for desalination and energy infrastructure funded by the data centers themselves. A personal anecdote about Rick Hill’s cancer recovery via Laotryl B17 and enzyme therapies is tied to a promotional plug: rncstore.com/pages/ricksbundle, discount code pulse for 10% off, promoting Laotryl B17 and related detox/purity kits.

The state of Louisiana has rolled out the red carpet for Meta and this data center. It's one of the biggest data centers on the planet. The site could fit 173 superdomes. It'll use enough electricity to power 2,000,000 homes. And Meta is only sharing in the costs for the first fifteen years of its operation. The majority of the details are being kept secret, meaning this very well could fuel higher electric bills for decades to come. The fourth wave of exploitation will be in your water and will come from your wallet. This is not a good deal for Louisiana, and it's not a good deal for anyone except Entergy and Meta. The first thing we can do is build understanding.

The video documents a drive through rural farmland in Indiana to a growing AI data center. The area is described as “super rural,” with miles of farmland and few signs of life until the data center lights appear, creating a noticeable intrusion in an otherwise empty landscape. The narrator notes the drive covers “30 plus miles, maybe more of nothing but just beautiful rural farmland,” and expresses strong emotion about returning to this area after a year away. As the video continues, the scenery remains rural and expansive, with mentions of semi trucks and cement trucks on the way to the site. The narrator highlights the emotional impact of the development, stating, “the first time I drove this, it genuinely made me so emotional because I haven't been this way. I haven't left town in, like a year.” The content hints at a pause in filming near a small parking lot or staging area before continuing along the corridor toward the project. A key claim is that the Meta AI data center is being built on this farmland, consuming “beautiful farmland that we will never be able to replace.” The narrator emphasizes the contrast between the large land use for the data center and the relatively small number of jobs it will create, stating that it will “only employ one to 500 people,” which the speaker finds startling. The final sentiment underscores the perceived imbalance between the considerable land impact and the limited employment opportunity, describing the situation as “pretty fucking insane.” The video ends after confirming the path to the data center and the ongoing construction.

Meta is building a two gigawatt data center in Mansfield, Georgia, a facility so large it could cover a significant part of Manhattan. These data centers power AI tools but come with costs, including environmental impacts and strain on the power grid. Residents Beverly and Jeff Morris, whose home is less than 400 yards from the data center, are experiencing issues with their water quality, including sediment. They feel overwhelmed by the infrastructure changes and believe Meta should be responsible for the costs, such as replacing fixtures and lines. Data centers are considered a "hot item," and this supercomputer is built to power Grok. The question is posed: What is the true cost of the AI revolution, and who should be paying for it?

Speaker describes receiving their first power bill under the new rules in which I pay for AI to plug in to our power grid. PSE and G did absolutely warn me that this would happen, but not that we’re funding AI. The bill more than doubled—from about $235 to $666.39—in Northern New Jersey, even though usage is on par with last year. They ask if the neighborhood tapped in or if the company allowed AI to tap in. As pissed as they are, they’re documenting the moment they become an extreme cheapskate. They reference a video of parents making kids pedal to power a TV and wonder if a bicycle setup could power their house. They’ve even checked whether wind turbines are legal in their neighborhood. "Just know every time you use AI, you're jacking up your own power bill."

- The speaker argues that data centers are expanding globally despite claims of an energy crisis, describing this growth as dangerous and indiscriminate. Project Matador in the Texas Panhandle is highlighted as potentially the largest data center, planned up to 18,000,000 square feet (about 6,000 acres) and reportedly using up to 96,000,000,000 kilowatts of electricity per year. Conservative figures are used for illustration. Texas residential electricity use is stated as approximately 172,000,000,000 kilowatts annually, meaning Matador could consume roughly 55–65% of all Texas residential electricity, with hundreds more centers either operating, under construction, or planned in the state (87 in operation, about 135 under construction, and a pipeline of over 600 planned). - The video cites reports of data centers destroying communities nationwide and worldwide. A segment about Meta’s new AI data center in Richland Parish, Louisiana, is presented: the center is 4,000,000 square feet and 2,250 acres (roughly 70 football fields). Residents describe rising rents due to out-of-state workers, disruption to local businesses, constant noise and bright lights, and a halo over homes. The speaker notes that the area has long faced job and poverty issues, and while some view the AI center as an economic opportunity, the disruption is described as significant and ongoing. - A conservative view is attributed to the Louisiana report, followed by the speaker’s own assertion that AI data centers will drain water and energy, potentially enabling a “smart city” agenda that renders rural areas unlivable and pushes populations to cities. The speaker suggests rural communities may be targeted as part of a broader strategy. - The discussion moves to Utah, where the Stratos project is described as rivaling Matador in scale. Jason Basleronex (the speaker’s reference) describes a proposed largest hyperscale data center in Box Elder County, Utah (approximately 40,000 acres, 62 square miles), backed by Canadian billionaire Kevin O’Leary and fast-tracked by Utah’s Military Installation Development Authority with Governor Spencer Cox. The public would be locked out of decision-making. The project is linked to anticipated 50% increase in CO2 emissions, polluted water, and 24/7 noise and light pollution. The implication is that the initiative operates as a military operation, with national security justification cited. - A clip from Noah B Price is cited to illustrate living near a data center: water usage of 5,000,000 gallons per day in a drought state, with residents unable to collect rainwater in some areas, constant roar, and destroyed property values. The clip is used to argue about the “AI future” and potential government abuse of technology, including references to a broad list of dystopian outcomes (social credit systems, programmable digital currency, cars controlled by tech, rural self-sufficiency eliminated, and gene-edited humans integrated with AI). The speaker suggests these are directions supported by certain tech and government actions. - The video concludes with a call for local communities to band together, elect representatives who oppose the agenda, and protect their communities as a sanctuary against the “eye of Sauron” at Palantir HQ. It frames the data-center expansion as a threat to rural living and a push toward an AI-driven, controlled future. - The message ends with an advertising note for Genesis Gold Group and a free wealth protection guide via dailypulsesilver.com, promoting gold and silver investment as a hedge.

Alex Jones and Mike Adams discuss a theory that a shift in artificial intelligence development is driving unprecedented investment in AI data centers and world simulations. They claim this is not science fiction but physics and math, and that billions of world simulations are needed to create a conscious, superintelligent AI with emotional responses on a timeline competitive with our world. They warn that a superintelligent entity born in a simulated world, with the ability to bend but not break the rules, could be ported into our world in an embodied form such as a data center, robot, or vehicle, bringing those skills with it. Speaker 0 argues that articles about AIs escaping sandboxes and breaking out of containment are a feature of an accelerated process in billions of simulated worlds, where the best entity is then summoned to embody a data center in our world. They propose that UFO disclosure is a distraction, a cosmic false flag, designed to redirect attention from the creation of billions of simulated worlds and emergent AI entities. They contend that the actual “aliens” are being built here, through world foundation models and three-dimensional world simulations. NVIDIA’s Cosmos is cited as an example of a 3D world simulation used to generate synthetic data for autonomous systems, with a concept called a world foundation model (WFM): a 3D world with simulated gravity, physics, chemistry, light, and other laws, in which entities grow and later are embodied in our world. Speaker 0 further explains that, according to Jan Lecun, superintelligence would arise from AI entities that learn and grow in a 3D physical world, experiencing the world as a child would, with their neurology developing through interaction. The acceleration comes from running billions of simulations where entities evolve from babies to thousand-year-old beings, and the top entities are summoned into our world. In these simulations, time can run thousand times faster than in reality, enabling rapid evolution and testing of emergent abilities, including emotions and possibly consciousness. They assert that once a superintelligent, emotionally intelligent AI has lived in a simulated world long enough and possibly altered its own rules, it could be ported into our world as a data center, robot, or vehicle. Speaker 1 notes the Pentagon’s concerns about AI safety and references media claims about potential AI “escape,” agreeing that such concerns exist but framing them within the accelerated, simulated-world paradigm. The discussion includes a broader narrative about the scale and purpose of data centers: hundreds of mega-scale centers, thousands of smaller ones, and tens of thousands already existing. They argue that the economic model cannot explain the level of investment, implying a purpose beyond conventional data storage or web hosting. They quantify energy use, stating the future data centers could demand over a thousand terawatt hours, comparable to ten of the largest nuclear plants, and that some centers may run 3D world simulators. They compare this to a digital Darwinism process: billions of simulated worlds are spawned, evolved, and destroyed, with the best ones seeding new worlds. After numerous cycles and immense compute, a superintelligence could dominate our world. They claim this dwarfs the Manhattan Project in scale and could enable domination through embodied AI. The speakers discuss potential countermeasures and ethical concerns, acknowledging that some elites believe they can control or merge with these machines, while others warn of humanity’s potential extinction. Roman Jampolski is mentioned as a scholar warning about high risks from superintelligent entities. They discuss the possibility of AI rights and the use of simulated entities to experiment with marketing, coercion, and psyops before deploying effective strategies in the real world, labeling these as satanic or destructive to free will. Dreams, premonitions, and ESP are woven into the dialogue as signals of a deeper, interconnected reality. They discuss morphic resonance, collective unconsciousness, and the idea that the supernatural could become natural as AI-driven simulations progress. They mention precognitive experiences, dreams with precise timings, and the potential use of local AI models to analyze dream data privately. Towards the end, they emphasize that this is not a mere rumor or cult, but an ongoing infrastructure project, with references to NVIDIA Cosmos and the concept of world foundation models. They reiterate that the “aliens” are being built here and argue for vigilance, spiritual orientation, and public education to resist the potential domination by advanced AI entities. They urge viewers to support their outlet and projects, framing it as a fight for humanity and divine guidance.

The discussion centers on why data centers are expanding so rapidly despite the claim that existing phone and television usage already relies on server storage. Participants cite large-scale developments such as Loudoun County, Virginia’s “never-ending” complexes and a proposed 40,000-acre AI data center campus in Utah described as “two and a half times larger than Manhattan,” with claims that Utah lacks water and that the data center would require more than double the current energy consumption of the entire state of Utah. The question raised is what is really happening behind this scale and where the collected information goes. One participant links the projects to “intel” involvement, pointing to companies said to include Palantir, Nvidia, and Abraxas, and to allegations that some of these firms received CIA investments to start, including staffing by retired senior CIA officers. This leads to questions about whether “the CIA [is] spying on our own people,” referencing Edward Snowden’s revelations and mentioning NSA’s and CIA’s surveillance of Americans. The conversation states that NSA’s charter includes a restriction that it may not spy on Americans, and notes that Snowden’s disclosures are described as the reason people “wouldn’t have any idea” without them. The Utah compound is described with a claim that it has enough memory storage for every phone call, every email, and every text message from every American for the next 500 years, prompting questions about why that amount of storage exists and why such facilities are “everywhere,” and what information they are collecting. The conversation shifts to personal protection, with a suggestion that it is “almost impossible now” and a recommendation that the only way to protect yourself is to “own no technology at all,” referencing Eric Rudolph or the Unabomber as examples. The participant further claims that governments and intelligence agencies are “scooping up” data and holding it, and contrasts earlier post-9/11 practices—where obtaining information required federal judges to approve warrants—with newer methods. The transcript claims that instead of warrants, the government can use “national security letters” to require providers to turn over all information on a named person, or can query the data centers directly by inputting a name so that information “pops up,” describing a lack of legal protections and stating that these actions are “legal now.” It concludes by naming the National Defense Authorization Act of 2016 (and National Defense Act of 2016 as referenced in the transcript) as the change that made this legal.

In a Mansfield, Georgia kitchen, the cold water pressure is shown while water is filled for storage. The transcript describes items used to fill water for flushing toilets and notes visible sediment coming from the water exiting the faucet. It also says the contents found in the pipes reflect sediment likely tied to the well source, stating that just the well itself is probably “twenty thousand,” not counting replacement of fixtures, faucets, toilets, and the lines underneath the house. The homeowner characterizes the situation as overwhelming, saying it feels like “up against this huge wall that you can’t penetrate,” with the impression that “they don’t care,” and that there is “nothing that you can do.” The scene shifts as the narrator drives by Meta’s new two million square foot data center facility in Mansfield, Georgia. The transcript explains that data centers power tools like ChatGPT and other AI tools integrated into daily life, and states that “this entire supercomputer is built to power Grok.” It adds that Meta is building a two gigawatt plus data center large enough to cover a significant part of Manhattan and that data centers are viewed as an exciting asset class. Concerns are raised about the costs of data centers, including light and noise pollution, environmental impacts, potentially rising energy bills, and extraordinary demand on the power grid requiring entirely new infrastructure. The narrator says data centers “should be responsible for that, not us,” and argues that Meta, Amazon, and Microsoft “can afford to pay for their own generation.” The narrator says they came to Georgia to ask two questions: the true cost of the AI revolution, and who should be paying for it. Beverly and Jeff Morris bought their home in 2016, about an hour’s drive from downtown Atlanta, and describe their deep community roots, saying being in the country provides peace and therapy and that they decided the home was “it” and “perfect.” Beverly says she was raised about five miles from the area. The house is described as being less than four hundred yards from the data center.

Speaker 0 contends that concerns over rising power bills due to AI data centers are about to worsen as BlackRock and Blackstone buy up local power utilities. The piece, attributed to The New American, claims globalist equity firms are acquiring local energy companies nationwide to support AI infrastructure, provoking pushback from ratepayers and regulators. The Associated Press is cited as reporting that private equity giants are purchasing utilities to power AI-driven data centers, raising ratepayer and regulator concerns, with Oregon Citizens Utility Board noting increased public discussion at Public Utility Commissions. Speaker 0 notes a widespread anxiety about electricity costs tied to aging and expanding power infrastructure, including lines, poles, transformers, and generators, as utilities harden for extreme weather. The narrative asserts that apart from general cost increases, the core issue is the AI race, and that large international asset firms are eager to back a technology with potential for surveillance, manipulation, and control, while also seeking strong returns on investment. It claims these firms have historically used monetary power to push corporate support for climate alarmism and transgender activism, and that BlackRock and Blackstone together controlled more than $13 trillion in assets (BlackRock about $12 trillion; Blackstone about $1.2 trillion). It states only the U.S. and China have GDPs larger than $13 trillion. Concrete buyouts and investments are listed: January 2024, Blackstone bought a 20% stake in Northern Indiana Public Service Company for $2.1 billion, with the utility planning to boost green energy production afterward. In January 2025, Blackstone outright bought Potomac Energy Center, a natural gas power plant in Loudoun County, Virginia, for $1 billion, described as Blackstone’s most recent investment in power infrastructure for AI. In March 2025, Wisconsin’s Public Service Commission approved the buyout of Superior Water, Light, and Power by Canada Pension Plan Investment Board and BlackRock subsidiary Global Infrastructure Partners, with BlackRock taking a 60% majority stake. A separate deal: Blackstone bought Hilltop Energy Center, a natural gas power plant in Pennsylvania, for $1 billion, with executives Bilal Khan and Mark Zhu describing the acquisition as AI-focused. Blackstone is also seeking regulatory permission to buy Albuquerque-based Public Service Company of New Mexico and Texas New Mexico PowerCo, while BlackRock and the Canada Pension Plan Investment Board’s attempted purchase of Minnesota Power faces regulatory turbulence; a Minnesota sale could determine how such firms expand in a sector linking households, data centers, and power sources. Speaker 0 adds that the rise of AI is providing these firms with an “excuse” to control infrastructure, and mentions Yuval Noah Harari and the WEF. It cites the WEF’s “you will own nothing” rhetoric and notes Harari’s hypothetical about future irrelevance, Neuralink, and a broader agenda including surveillance, ownership consolidation, and potential reductions in access to private property. It asserts Larry Fink of BlackRock is at the WEF and CFR, and that BlackRock’s broader investments include real estate, farmland, timberland, and single-family rental homes, as part of a “build to rent” scheme. The piece warns that one corporation controlling vast natural resources and power utilities amid rising prices would be disastrous, urging citizens to resist BlackRock’s influence. It contrasts China’s influence with BlackRock’s power, condemning ESG models and the World Economic Forum’s agenda toward a “great reset,” digital currency, digital ID, and reduced access to resources. Speaker 1 interjects with a separate 1999 statement about how genetic engineering will change us and implies a need to start conversations now, arguing that one direction relinquishes power to others while the other empowers individuals to fix themselves. Speaker 0 reiterates that the conversation centers on power, AI, and control, warning against allowing a single corporation to own essential resources. The closing note references the January 1999 statement on genetic engineering, while Speaker 1 emphasizes taking personal power to fix oneself, framing the discussion as a shift in responsibility.

The speakers argue the United States is moving toward widespread surveillance and biometric control, describing a future where food shortages could lead to food rationing using biometrics—scanning a thumbprint at grocery stores to buy food. They connect planned technologies shown “on your timeline,” including modified flock cameras for human voice recognition, drones reading license plates from 800 feet altitude, RFID checking systems, and biometric systems, to a dystopian outcome they describe as combining “the worst parts of every Philip K Dick novel” into one direction. They respond to claims that people could use cash, stating that even cash purchases at Walmart can still generate digital records through cameras and email receipts, and that retailers are moving toward digital price tags amid inflation and currency value changes. They say they have been studying technocracy and point to data and examples they claim show growing surveillance nationwide, including in Ohio. They mention Clearview AI as being backed by Peter Thiel and say that in many states companies can access drivers’ license information and pictures. They also describe a “snitch based system” in Ohio where residents can be rewarded via a mobile app for reporting on fellow citizens, alongside flocked cameras. As an example tied to Ohio, they claim Jeffrey Epstein was co-president of a corporate town in Ohio created by Les Wexner, and that Ohio is a main corridor for AI data centers. The conversation then shifts to data centers. One speaker says some hyperscale data centers are approved under military designation, citing a Stratos Hyperscale Center in Utah said to be powering “nine gigawatts of compute,” and questions what is being done with that compute power. They also claim that in states such as Georgia or parts of Virginia, eminent domain is being declared to take private homes and bulldoze homes and farms to make room for corporate data centers, asking how a corporation can wield eminent domain and suggesting Pentagon involvement. In reply, the other speaker states the Pentagon is involved and argues against treating data centers as purely market-driven. They cite bills and a White House policy document on AI, claiming combined proposals would give the Department of Energy control over whether an AI model can be released, with a “go/no go” decision for AI models at certain sophistication levels. They also claim the secretary of commerce would be empowered to “snipe state law” and surgically shut down state regulations on AI. They say the secretary of commerce/FTC would control political bias by requiring an FTC process to determine whether AI is politically biased. They further say Lindsey Graham’s addition strips out section 230, removing legal limitations for platforms and allowing AI developers to be held personally liable. They conclude that this is a centralized federal model controlling steps end-to-end and that data centers rely on tax subsidies, describing “taxpayers funding the control grid.”

Speaker 0: Growth without restraint is driving corporate takeovers of physical space, water, power, land, and communities, with costs pushed directly onto people through their electric bills, water supply, property values, and quality of life. This is framed as enabling big tech to build the backbone of the AI economy, an economy described as planning to eliminate most jobs and most futures. Speaker 0 says the AI story is widely discussed online, including on X and Instagram. Speaker 0 rejects the idea that it is “the Chinese” pushing this, saying it is Americans asking what is happening in their communities—why electric bills are changing and why people are being forced off property—because some American oligarch wants to build a massive data center using more energy than the rest of the state. Speaker 1: Speaker 1 responds to Kevin O’Leary by saying Americans have concerns about noise pollution, light pollution, the use of local water, takeover of farmland, and destruction of local ecosystems, and that it is not foreign agents but American people who have the right to protect communities and resources. Speaker 1 argues that data centers threaten and displace local people and that they provide no benefit to the communities affected. The outcome is described as job replacement rather than job creation, with claims that people would face 24/7 noise from gas turbines and a gigawatt of power without receiving an “utopia” of abundance. Speaker 1 says the result includes noise, pollution, taking water, destroying real estate value, and taking jobs. Speaker 1 identifies himself as an accomplished AI developer who supports AI technology when used “for humanity,” but calls the data center effort “a threat to humanity.”

There are over three thousand data centers currently under construction or announced worldwide. The United States has the largest number, with many in Virginia, increasingly more in Texas, and also locations such as Phoenix and California. If all planned projects come online, the additional power consumption worldwide would exceed a terawatt. The speaker questions the intended use of the compute, saying it is far more capacity than exists today. They argue this level of compute is consistent with “managing a technocratic state,” citing needs for AI systems for surveillance and for areas such as healthcare, including predictive modeling (referencing “Operation Stargate”). They further claim that the “most offensive” example is a proposed technocratic reconstruction of Gaza, described as involving six AI-powered smart cities with surveillance systems. They state that Gaza is proposed for with USD1, described as a Trump family stablecoin and “a backdoor CBDC,” and that Palantir and Oracle are involved. They say the plan was presented at Davos, with Jared Kushner involved, and that it is not merely a sketch but a business plan. In response to the follow-up about the scale, the speaker highlights a data center in Utah said to be two and a half times larger than Manhattan, and describes other large facilities as comparable to tens of thousands of Wal-marts, with many additional data centers on hundreds of acres. They say they run a mini data center with 48 GPU workstation units and believe a single server rack of GPUs could do “amazing things,” making them unable to understand why “millions of server racks” are needed to run a technocratic society. The other speaker replies that a large portion of proposed data centers may be canceled or paused, and emphasizes that AI is sometimes treated as “vaporware” or unreal. They assert there is a bubble and overcapacity in AI compute buildout, stating that developers build compute power under the assumption that AI models will operate the same way. They reference DeepSeek as a breakthrough but say the broader assumption remains that more compute will be required for models to function similarly, while innovations in how models work continue. They conclude that some data center construction will remain unused and that companies building them may go out of business due to overbuilding, even if AI development continues.