Note: This is a research note supplementing the book Unscarcity, now available for purchase. These notes expand on concepts from the main text. Start here or get the book.

Compute Clusters: The Factories of the Intelligence Age

How AI training facilities became the oil refineries of the 21st century - and why your grandchildren might receive a compute allocation alongside their birth certificate.

What Is a Compute Cluster, Really?

“Compute cluster” sounds impressively technical but obscures something simple. A compute cluster is just a lot of computers working together on the same problem. The magic is in the “working together” part.

When you train a large language model like GPT-4 or Llama 4, the task is far too massive for any single computer, even an obscenely powerful one. You need to split the work across thousands of processors that constantly communicate, share results, and coordinate their efforts. If one machine calculates something, every other machine needs to know about it immediately. The lag between them must be measured in microseconds, not seconds. Clusters like this one are the first stop on a longer journey: when intelligence gets cheap, value doesn’t vanish, it migrates to whichever layer is still scarce, and right now that’s the compute itself.

Building a pyramid is a useful analogy. One person with a wheelbarrow would take millennia. A thousand people need to coordinate: who’s carrying what, where’s the next block going, don’t drop that on Steve. A compute cluster is that coordination problem, solved at the speed of light.

The processors doing this work are Graphics Processing Units (GPUs), originally designed to render video game graphics but repurposed for AI because they excel at doing many simple calculations simultaneously. If a traditional CPU is a brilliant surgeon performing delicate operations one at a time, a GPU is a factory floor with thousands of workers each doing one small task very quickly.

The Current State of Play: Welcome to the GPU Arms Race

In 2024-2025, we crossed into new territory. The compute clusters training frontier AI models became genuinely staggering in scale:

Colossus: From 100,000 GPUs to a Gigawatt Landlord

Elon Musk’s AI company, xAI, built what was briefly the world’s largest AI supercomputer in a converted Electrolux factory in Memphis, Tennessee. The timeline was absurd: from decision to operational in 122 days, when similar projects typically take years. By September 2024, it was running 100,000 Nvidia H100 GPUs.

Then they doubled it, in 92 days, and kept going.

Colossus 2 went up faster still: a first cluster of about 110,000 GB200 processors and 210 megawatts in 91 days, then a second of 110,000 GB300s and 220 megawatts in 64, according to the IPO prospectus of SpaceX, which merged with xAI in February 2026. On its first earnings call, on August 4, SpaceX said it ended June with 1.4 gigawatts of nameplate compute, up from 400 megawatts a year earlier, and expects more than 2 gigawatts by the end of the year.

How many chips that is depends on who is counting. In September 2026 Musk posted a tally on X that adds up to 230,000 GPUs at Colossus 1 and 550,000 at Colossus 2, with up to 660,000 more promised by late December, the last third only “if we get lucky.” Epoch AI, which estimates capacity independently from satellite imagery, credits Colossus 2 with 440,000 chips and 946 megawatts of IT power on imagery last taken in June, and ranks it the largest known AI data center in the world by IT power. Three more data centers are planned nearby, one of them nicknamed “Macrohardrr.”

Then came the twist: the fortress became a rental property. SpaceX went public on June 12, 2026 at $1.77 trillion, the largest IPO on record, and its filings revealed that the biggest users of Colossus are xAI’s competitors. Anthropic pays $1.25 billion a month through May 2029 for roughly 325,000 GPUs across both sites (early reports had it leasing only Colossus 1). Google pays $920 million a month from October 2026 through June 2029 for about 110,000 GPUs, which a Google Cloud spokesperson described as “bridge capacity.” Reflection AI signed on at $150 million a month, and in September SpaceX’s CFO disclosed a fourth, unnamed customer at about $1.11 billion a month from December 1. Once all four are paying, that is $3.43 billion a month, about $41 billion a year, most of it from companies SpaceX names as competitors, and the three disclosed contracts can each be cancelled on 90 days’ notice. Anthropic’s draft IPO prospectus, reported by Reuters on September 29, gives the tenant’s view: agreements with SpaceX’s AI unit that could reach $84.5 billion through 2029, nearly double what the May deal implies. The builders are becoming landlords, and the tenants are giants.

This is no R&D project. It is industrialized intelligence production at a scale that didn’t exist two years ago, and it draws on a shared commons. Colossus 2 runs on gas turbines across the state line in Mississippi. After a Clean Air Act complaint from the NAACP, SpaceX agreed with state regulators to remove all 69 temporary turbines by July 2027 and replace them with a permanent 1.2-gigawatt plant. Memphis is weighing a 12-month moratorium on new data centers, one of many local vetoes on the AI buildout: the council passed it on first reading in August, pulled it from the September 15 agenda, and takes it up again on October 6. The electrons flowing into a private training cluster are electrons not flowing into homes, hospitals, and schools.

Meta’s Infrastructure: Millions of GPUs

Meta started the AI boom by accumulating 350,000 H100 GPUs by end of 2024. That was just the beginning. In early 2026, Meta and Nvidia announced a multiyear partnership covering “millions” of Nvidia’s Blackwell and upcoming Rubin GPUs, one of the largest single infrastructure commitments in semiconductor history. A week later, Meta struck a separate deal with AMD for up to 6 gigawatts’ worth of GPUs. The company has since raised its 2026 capital expenditure guidance to $130 to $145 billion, roughly double the $72.2 billion it spent in 2025. Meta is also building Hyperion, an AI campus in Louisiana that it plans to expand to 5 gigawatts of compute capacity, fed by ten gas-fired plants with 7.5 gigawatts of capacity, seven of which still await a regulatory vote. And on July 1, 2026, Bloomberg reported that Meta is building a cloud business to sell its excess AI capacity, reportedly called Meta Compute, offering both raw GPU cycles and hosted model access in direct competition with AWS, Azure, and Google Cloud. Wall Street’s verdict: the stock jumped about 10% in a day.

Four weeks later the market priced the same strategy in reverse. Meta’s second-quarter results, reported July 29, 2026, showed $31.1 billion spent on data centers, servers, and chips in a single quarter, up 83% year over year, against $31.9 billion of operating cash flow. Free cash flow came in at $784 million, down from $8.55 billion in the same quarter a year earlier, and the stock fell as much as 10% after hours. On the call, Zuckerberg said Meta is “getting a lot of offers for compute at a significant premium over what we paid for it.” He did not rule out taking them, but said “it would be foolish to basically just sell all of the compute and take a short-term profit” when intelligence built on it is worth a multiple. Sit with what that premium implies. Buyers will pay above cost for someone else’s already-energized capacity, because a chip order clears in months while a gigawatt-scale grid interconnection takes years. Money can buy silicon. It cannot buy a place in the queue.

Their Llama 3 model trained on a cluster of 16,384 H100s for 54 days, during which they experienced 148 interruptions from faulty GPUs and 72 from memory failures. This is rocket science that involves replacing engines mid-flight.

Oracle’s Zettascale Vision

In September 2024, Oracle began taking orders for what it called the first “zettascale” AI supercomputer: up to 131,072 GPUs in a single cluster, more than three times as many as the Frontier supercomputer at Oak Ridge. Two years later that ceiling looks modest: Oracle says it delivered more than 300,000 GPUs to AI customers in a single quarter.

The Stargate Project: $500 Billion Meets Reality

Announced in January 2025 with a White House press conference, the Stargate Project represents perhaps the most ambitious AI infrastructure buildout ever conceived. SoftBank, OpenAI, Oracle, and MGX pledged up to $500 billion over four years, with $100 billion deployed immediately.

By late 2026, reality has set in. The flagship site in Abilene, Texas has had its troubles, including a severe weather event that, in Sam Altman’s words, temporarily “brought things down.” It is also delivering. Oracle told investors in September that it handed over 131,000 GPUs at Abilene in one quarter and has delivered six of the campus’s eight buildings, 618 megawatts. (Epoch AI’s satellite estimate is lower: 421 megawatts of IT power.) Microsoft has taken two further buildings next door, bringing the complex to ten buildings with roughly 2.1 gigawatts of projected capacity.

The map keeps growing. OpenAI, which in March owned no data centers at all, announced in July that it is designing and developing its own in Effingham County, Georgia, with 3.2 gigawatts of power to arrive in phases between 2028 and 2032. In August, Nvidia agreed to guarantee up to $105 billion of lease obligations on an OpenAI campus in Ohio, about 4.25 gigawatts under 20-year leases. The chip supplier is co-signing its customer’s rent, which is what compute as collateral looks like.

So does the list of things that can stop it. Oracle sent a force majeure notice to the developer of Project Jupiter, a 2.45-gigawatt Stargate campus in New Mexico, that would let it delay payments if the site misses its 2028 target; Oracle says the project “remains on our planned schedule.” A week later floods reached the campus; Oracle said no critical infrastructure was damaged. In Texas, Governor Greg Abbott directed the state’s environmental agency on September 21 to pause all permits for data center projects until the grid operator completes its review.

What’s Inside These Clusters: The Hardware

The Workhorse: Nvidia H100

The Nvidia H100 (codenamed “Hopper”) was the chip that defined the 2023-2024 AI boom and remains widely deployed. Key specifications:

80GB of HBM3 memory (the fastest memory type available at launch)
3.35 terabytes per second memory bandwidth (how fast data moves)
4 petaflops of AI compute (a petaflop is a quadrillion calculations per second)
~$25,000-40,000 per chip when new

The H100 was genuinely scarce in 2023-2024, with AI companies hoarding them like gold bars. Rental prices have come down since, but the old chip refuses to depreciate on schedule. Silicon Data’s index reads $2.53 per GPU-hour; Ornn’s showed $3.28 on September 7, 2026, up 22% in a month.

Blackwell: The Current Generation

Nvidia’s Blackwell architecture shipped at scale through 2025 and drove Nvidia to a record $215.9 billion in fiscal year 2026 revenue. In the quarter that ended July 26, 2026, Nvidia reported revenue of $96.2 billion, up 106% from a year earlier. The B200 GPU represents a generational leap over Hopper:

192GB of HBM3e memory (2.4x the H100)
8 terabytes per second bandwidth (2.4x the H100)
20 petaflops of AI compute (5x the H100)
~$30,000-40,000 per chip

The GB200 NVL72 system connects 72 Blackwell GPUs to act as a single massive processor with 1.4 exaflops of AI performance. That’s 1.4 quintillion calculations per second. For perspective, the human brain is estimated to perform about one exaflop, so this rack of GPUs matches 1.4 human brains in raw computational throughput.

The Next Wave: Vera Rubin

Vera Rubin, the successor to Blackwell, began production shipments in August 2026, and Nvidia expects it to account for about 20% of data center revenue in the following quarter. The Vera Rubin NVL72 rack links 72 Rubin GPUs with 36 Vera CPUs using NVLink 6 switches. Nvidia’s claim for it is 30x higher throughput per megawatt and 35x lower token cost than Grace Blackwell Ultra. At GTC in March 2026, Jensen Huang said he expects purchase orders for Blackwell and Vera Rubin to reach $1 trillion through 2027, double the $500 billion the company had projected a year earlier. Nvidia expects supply “to remain a bottleneck at least through the end of fiscal year 2028.” The next architecture after Rubin, codenamed Feynman, is already in development targeting TSMC’s A16 process node.

The Networking: NVLink and InfiniBand

Raw GPU power means nothing if the chips can’t talk to each other fast enough.

NVLink is Nvidia’s proprietary interconnect for GPU-to-GPU communication within a single server or rack:

1.8 terabytes per second of bidirectional bandwidth (the latest version)
7x faster than PCIe Gen 5 (the standard computer connection)
Enables GPUs to share memory directly, as if they were one chip

InfiniBand connects servers across the data center:

400-800 Gb/s per port (with Quantum-X800 InfiniBand now shipping)
Under 100 nanoseconds latency (a nanosecond is a billionth of a second)
RDMA capability (Remote Direct Memory Access - GPUs can read each other’s memory without involving the CPU)

The typical architecture: NVLink connects GPUs within a node, InfiniBand connects nodes across the cluster. Together, they make thousands of GPUs behave like one giant processor.

The Power Problem: These Things Are Hungry

The compute cluster buildout collides with physical reality: these facilities consume obscene amounts of electricity.

According to the International Energy Agency:

Global data center electricity consumption was about 415 TWh in 2024 (~1.5% of global electricity)
U.S. data center consumption is projected to rise by around 240 TWh by 2030, up 130% from 2024
The agency’s 2026 update puts 2025 at 485 TWh, up 17% in a single year, and projects about 950 TWh by 2030, around 3% of global electricity demand

AI specifically consumes 10-20% of current data center energy, but that fraction is rising rapidly, potentially to 35-50% by 2030. Data center electricity demand is growing at roughly 15% per year, more than four times faster than all other sectors combined.

SpaceX is building its own 1.2-gigawatt gas plant for Colossus. Meta’s Hyperion campus in Louisiana is planned around 7.5 GW of gas-fired power, which would add more than 30% to the state’s entire grid capacity.

Google’s environmental report, published July 2, 2026, shows what this looks like on one company’s meter: electricity use up 37% in 2025, the largest single-year increase in its history, with data centers alone consuming 42 million MWh. That is one firm drawing roughly a mid-size country’s annual power. Google’s own explanation for why it expects the curve to steepen is the sentence planners should be reading twice: the buildout is accelerating faster than the grid is decarbonizing.

This is more than an engineering problem. It is a civilizational constraint.

The Fusion Connection

The tech industry’s obsession with fusion energy is grounded in arithmetic, not PR. When you’re planning to run data centers that individually consume gigawatts of power, you need energy sources that scale.

Microsoft has signed a deal with Helion for 50 megawatts of fusion power from a plant due to begin operating by the end of 2028. Helion broke ground in 2025 and has since softened what 2028 means: its website now promises only “initial operations,” and an executive told a fusion conference, “We will not, in no universe, have 50 megawatts hit full operation in 2028.” OpenAI’s own Helion deal, reported in March 2026 as 5 GW by 2030 and 50 GW by 2035, remains a negotiation; Helion says it “has not announced any new customer agreements.” Google signed a 200 MW fusion PPA with Commonwealth Fusion Systems, which has raised $4 billion in total and is targeting first plasma from its SPARC machine in 2027. The Fusion Industry Association counts $14.24 billion raised by fusion companies since 2021; an EU tally had already reached $15 billion by September 2025.

The timeline is aggressive but not arbitrary. Current grid infrastructure simply cannot support the projected AI data center buildout. Something has to give. Either AI development slows, or we find new energy sources.

This is the fuel component of the Unscarcity tripod (The Brain, The Body, The Fuel). Fusion and AI are symbiotic necessities, not merely adjacent technologies.

The Hyperscaler Investment: Hundreds of Billions

The capital flowing into AI infrastructure defies historical precedent:

Company	2026 Guidance: Previous → Latest	Focus
Amazon (AWS)	~$200B → ~$220B	Cloud AI infrastructure
Google	$180-190B → $195-205B	Cloud, Gemini training
Meta	$125-145B → $130-145B	Llama training, social AI
Microsoft	~$190B → ~$175B (lease accounting change)	Azure, OpenAI/Stargate
Oracle	$55.7B spent in FY2026 → $90-95B guided for FY2027	Stargate sites, cloud AI

In August 2026, Nvidia’s CFO told investors that capex by the top five hyperscalers is expected to reach nearly $800 billion in 2026 and $1.3 trillion in 2027. A month later Goldman Sachs strategists put 2027 at $1.2 trillion and estimated that the five would need about $300 billion a year in AI revenue to break even. The labs are signing for their share of the bill. Anthropic’s prospectus lists at least $518 billion of infrastructure commitments over the next decade, among them $161.2 billion in equipment leases from Broadcom, $111.1 billion to Google, $110 billion to Amazon, and $31.4 billion to Microsoft, and about 80% of the total is owed whether or not the capacity is used. By 2030, cumulative AI data center capital expenditure is projected to exceed $5.2 trillion.

These numbers are difficult to contextualize. The entire Apollo program cost about $280 billion in today’s dollars. The 2026 AI infrastructure spend alone is nearly three times that.

The Geopolitics: Compute Is the New Oil

Sam Altman has called compute “the currency of the future”, “possibly the most precious commodity in the world.” Jonathan Ross of Groq echoes the sentiment: “Compute is the new oil.”

The phrase is not hyperbole. The concentration is staggering:

The United States controls ~75% of the world’s AI supercomputing capacity
China holds ~15% (and falling, due to tightened export restrictions)
Nvidia commands ~85% of the GPU market overall, with 80%+ of AI training chips (down slightly from 92% in early 2025 as AMD reaches ~7%)
TSMC manufactures ~90% of the world’s advanced chips, and in July 2026 raised its capital budget for the year to $60-64 billion

The U.S.-China semiconductor conflict is, in effect, a war over the means of intelligence production. The CHIPS and Science Act allocated over $52 billion to incentivize domestic chip manufacturing. TSMC announced an additional $100 billion investment in Arizona in July 2026, which chairman C.C. Wei said would probably mean four more fabs. The explicit goal is to reduce dependence on Taiwan, which sits 100 miles from mainland China.

Export controls have made China a marginal producer of AI chips. They have also shut Nvidia out of China: the company’s latest quarterly filing says it was “effectively foreclosed from competing in China’s data center computing/compute market.” Estimates of Nvidia’s 2025 share of China’s AI accelerator market range from 55% to 40%; Bernstein sees 8% for 2026, and TrendForce expects Chinese-made chips to take 90% of the domestic market. DeepSeek founder Liang Wenfeng stated bluntly: “Money has never been the problem for us; bans on shipments of advanced chips are the problem.”

Yet China is responding. SMIC continues advancing domestic chip production. Beijing’s “Big Fund” pours billions into semiconductor development. A technological decoupling is underway, creating two increasingly incompatible AI ecosystems, American and Chinese, competing for global influence.

The Taiwan Question

The uncomfortable truth: if China blockades or invades Taiwan, the global tech economy collapses overnight. Modern militaries would face an immediate semiconductor drought. Every iPhone, every AI model, every advanced weapon system depends on chips that mostly come from one small island.

This is why TSMC is expanding globally: to Japan (with Japanese government subsidies, already upgrading its Kumamoto fab from 6nm to 3nm), to Arizona (where the additional $100 billion is headed), to Germany. Geographic diversification has become a national security imperative, well beyond ordinary business strategy.

Terafab: Owning the Whole Stack

The vertical-integration logic reaches its extreme with Terafab. Announced in March 2026 as a collaboration between SpaceX and Tesla, with Intel joining in April, it is a chip manufacturing initiative with a long-term goal of producing one terawatt of compute hardware a year. On August 6 the companies named the site and the opening bid: Grimes County, Texas, and an initial $16.8 billion for a plant of 100 million square feet. (A May filing had proposed $55 billion to start and $119 billion if every phase is built.) State agreements list a construction start of December 1 and completion by the end of 2028. The stated purpose is chips for Tesla’s Optimus robots and Cybercabs and for SpaceX’s space-based data centers.

The March announcement came with no timeline and no price; both now exist. The ambition remains the hard part. Building a leading-edge fab is among the hardest industrial undertakings humans attempt: TSMC spent more than $30 billion on its Arizona fabs and they still aren’t fully operational, and the entire global semiconductor industry currently produces chips consuming roughly 300-400 gigawatts at peak load. A single facility claiming a terawatt of yearly output would, in theory, produce more compute capacity per year than the world currently uses.

Whether or not Terafab is built on Musk’s stated timeline (his track record runs late but directionally correct), the structure is the point. One ecosystem would design the chips, manufacture them, install them in cars and robots and rockets, train AI on them, and deploy the result into the physical world, with no congressional oversight of the architectures, no independent audit of the models baked into the silicon, and no public review of the supply chain. Compute is becoming the new oil, and Terafab is an attempt to own the refinery, the pipeline, and the gas stations at once. The deeper concentration problem is taken up in Who Owns the Abundance Stack?, Infrastructure Libertarianism and The Iron Law of Oligarchy.

Why Individual Chips Don’t Matter (And Why Clusters Do)

A single H100 is impressive but useless for frontier AI training. The models are simply too large. GPT-4 is rumored to have around 1.7 trillion parameters. Llama 3’s largest version has 405 billion. Each parameter needs to be stored, updated, and communicated.

The math works like this:

A 1 trillion parameter model needs ~2 terabytes just to store the weights in half-precision floating point
During training, you also need to store gradients and optimizer states, roughly 16 bytes per parameter
That’s 16 terabytes of memory for a 1 trillion parameter model
A single H100 has 80GB of memory

You need at minimum 200 H100s just to hold the model, plus additional GPUs for batch processing. In practice, training requires thousands to tens of thousands of GPUs running for weeks or months.

This is why clusters matter. The core technological achievement is the ability to coordinate tens of thousands of processors on a single training run, with minimal communication latency, maximal bandwidth, and reliable fault tolerance. Any individual chip is impressive on its own, but the orchestra is what transforms the field.

The Coming Democratization (Maybe)

The Unscarcity thesis becomes concrete at this point.

Right now, frontier AI training is confined to about a dozen organizations worldwide: OpenAI, Anthropic, Google DeepMind, Meta, xAI, Mistral, a handful of Chinese labs, and some well-funded startups. The capital requirements (billions of dollars in GPUs, power, and engineering talent) create natural barriers.

But three forces could democratize access:

1. Inference Is Cheaper Than Training

Training GPT-4 required months on a massive cluster. Running inference (getting answers from the trained model) is far cheaper. Cloud providers now offer API access to frontier models for pennies per query. The capability is increasingly accessible even as the means of production remains concentrated.

2. Smaller Models Are Getting Better

Distillation techniques, efficient architectures, and better training data mean smaller models can approach the performance of larger ones. Llama 3’s 8 billion parameter model outperforms GPT-3.5 (175 billion parameters) on many benchmarks. The floor is rising.

3. Decentralized Compute Networks

Emerging networks like Aethir aggregate idle GPU capacity globally. The idea: if a million people each contribute a small amount of compute, you get a distributed supercomputer. The technology is nascent, and the concept is sound.

Connection to the Unscarcity Vision: Universal Basic Compute

This brings us to Universal Basic Compute (UBC): the idea that every citizen should receive a guaranteed allocation of AI processing capacity.

If compute is becoming the means of production, the modern equivalent of land in an agrarian economy, then distributing compute is distributing economic agency. A citizen with a compute allocation can:

Run personal AI agents that manage life administration
Participate in pooled cooperatives for larger projects
Delegate their allocation to Mission Guilds in exchange for services
Create, research, or build using the same tools as major corporations

The Foundation layer in the Unscarcity framework could eventually provision compute the way it provisions housing or food: as infrastructure for dignified existence, a baseline rather than a luxury.

None of this is fantasy. Research institutions already allocate compute quotas. National cloud initiatives in Saudi Arabia, UAE, and Singapore are building sovereign AI infrastructure. The concept of “compute as public utility” is emerging in policy discussions.

But achieving UBC requires:

Sufficient total compute - currently a constraint, but Blackwell, Vera Rubin, and future generations are multiplying capacity at 10x per generation
Distribution infrastructure - cloud platforms that can allocate and meter compute fairly
Accessible interfaces - so “using compute” becomes as intuitive as “using electricity”
Governance frameworks - to prevent concentration, corruption, and hoarding

The Diversity Guard mechanism addresses governance. The Civic Service program includes AI orchestration training. The EXIT Protocol negotiates with existing power holders for gradual transition.

What to Watch For

2026-2027

Vera Rubin ramps through a supply bottleneck Nvidia expects to last at least through its fiscal 2028; the Feynman architecture follows (2028)
Stargate’s first gigawatt of Nvidia systems, targeted for the second half of 2026, and whether New Mexico’s Project Jupiter holds its 2028 date
Hyperscaler CapEx, nearly $800 billion in 2026, heads for $1.2-1.3 trillion in 2027
The compute landlord market matures: Meta Compute, SpaceX-xAI leases, and CoreWeave compete to rent capacity - watch whether pricing trends toward commodity utility or gatekept rent
Two-ecosystem decoupling accelerates as Nvidia’s share of China’s AI chip market heads toward single digits
TSMC starts on its additional $100 billion in Arizona
States move from cost allocation to outright pauses: Memphis revisits its moratorium on October 6, 2026, and Texas regulators report back by October 19 (who pays for AI’s electricity)

2027-2030

First commercial fusion plants potentially come online (Helion’s “initial operations” in 2028, CFS/Google early 2030s)
Terafab’s planned completion and first capacity at OpenAI’s Georgia and Ohio campuses (2028)
AI-specific chip architectures (custom silicon from Meta, Amazon, Google) mature alongside Nvidia
Inference compute demand surpasses training compute
UBC pilot programs in selected jurisdictions

2030+

Fusion energy begins meaningfully powering data centers
Foundation-layer compute infrastructure in early deployments
Cumulative AI infrastructure spend exceeds $5 trillion
The contours of “who controls intelligence production” become clear

The Stakes

Something specific is being decided in the GPU warehouses of Memphis, Abilene, and undisclosed locations worldwide:

Who will control the means of intelligence production?

If the answer is “a handful of corporations and nation-states,” we get the Star Wars trajectory: technological feudalism with better graphics. The abundance AI creates concentrates at the top. Everyone else becomes economically irrelevant but biologically alive.

If the answer is “distributed infrastructure with universal access,” we get something closer to the Unscarcity vision, where compute is a public utility, AI capability is broadly held, and the Foundation layer provides dignified existence while the Frontier layer rewards contribution.

The race to build compute clusters is a commercial competition, but it is also the construction of the infrastructure that will determine how intelligence is distributed across human civilization.

The factories of the Intelligence Age are being built right now. The question is whether they’ll produce liberation or lock-in.

AI Compute Clusters 2026: The $800B GPU Arms Race