Elon's Anthropic Deal, The Next AI Monopoly?, "FDA for AI" Panic, Trading the AI Boom

The recent lease agreement between Anthropic and Elon Musk marks a major transformation in the AI infrastructure landscape. By leveraging SpaceX's Colossus data center and related assets, Musk has positioned X.ai and his broader ecosystem to directly compete with cloud giants while solving critical compute bottlenecks in the current AI boom.

Elon's Data Center and Energy Infrastructure Build-Out Advantages

Elon Secured Massive Computational Capacity Through Spacex's Colossus and Related Data Centers, Positioning X.ai to Compete Directly With Aws, Azure, and Google Cloud

Elon Musk’s foresight in building out data center and energy infrastructure at unprecedented speed has fundamentally changed the balance of power in AI compute. SpaceX’s Colossus—and companion facilities MacroHard and MacroHarder, which together boast over 1.2 gigawatts of capacity—are now online, enabling X.ai to trade and train AI models at scale. This build-out vaults X.ai’s Elon Web Services (EWS) into the realm of hyperscalers, placing them shoulder-to-shoulder against established titans like Amazon Web Services, Google Cloud, and Azure. As Brad Gerstner describes, SpaceX now boasts a “five layer cake” stack: launch, connectivity, compute hyperscalers, space data centers, applications and models, and other bets. This vertical integration across space, energy, and compute positions Elon as a kingmaker in AI infrastructure.

X.ai's Elon Web Services (Ews) Projected to Generate $4-5 Billion Annually, Subsidizing Grok's Development and Easing Financial Pressure

The emergence of EWS brings a new high-margin revenue stream, projected by Gerstner to generate an additional $4 to $5 billion a year—an amount that materially offsets the substantial investments required to train X.ai’s Grok and build out further capacity. This incremental revenue, above already robust analyst estimates, means X.ai can sustain aggressive R&D and infrastructure investments without immediate pressure for commercial returns. Chamath Palihapitiya notes that this terrestrial capacity provides a structural core business for Elon, blunting valuation anxieties tied solely to future and riskier orbital data centers.

Elon Recognizes Power and Compute As Ai Bottlenecks, Builds Data Centers Faster Than Competitors

Elon saw the growing bottleneck in power and compute before most of the industry and executed rapidly to secure both. As a result, he has become the rare operator with the scale, energy assets, and infrastructure to support both his companies’ needs and those of the broader AI ecosystem. Gerstner underscores Elon’s unmatched ability to “convert electrons to tokens,” connecting expertise in battery deployment, solar, and gigafactories from Tesla and SolarCity with hyperscale compute operations.

Anthropic-Elon Lease Agreement Solves Compute Constraints

Anthropic, Despite Data Center Limits, Gains 220,000+ Nvidia Gpus and 300+ Megawatts, Boosting Api Volumes and Removing Rate Limits

Anthropic, like many AI companies, has faced severe compute and power constraints. The new lease agreement provides Anthropic access to over 220,000 Nvidia GPUs and more than 300 megawatts of energy, via Colossus. This immediately lifts prior restrictions: Claude users no longer face API rate limits or peak usage caps, and paid Opus API volumes have dramatically increased. Prior to this, Anthropic’s scaling and revenue growth were not constrained by demand but by limited supply of high-performance compute—especially power.

Elon Monetizes Colossus By Leasing Spare Capacity to Anthropic, Transforming Infrastructure Costs Into Revenue While Reserving Resources For X.ai's Model Training

The leasing of spare Colossus capacity to Anthropic is a pivotal financial and strategic move. The “less connected, H100s, great for inference” assets are now monetized in a big way, alleviating financial pressure on X.ai, allowing it to invest in Grok’s development, and providing a profitable alternative to [restricted term] lying idle. This model lets X.ai remain a frontier AI lab without the burden of massive, unprofitable upfront capital commitments. Importantly, Elon reserves enough resources for X.ai’s needs, ensuring internal model training is never neglected.

Collaborative Ai Development: Frontier Labs Partner on Infrastructure, Compete On Models, Contrasting With Zero-Sum Competitive Dynamics

A noteworthy theme is the cooperative dynamic emerging among AI labs. While Anthropic and X.ai remain fierce competitors in AI models, the infrastructure layer sees collaboration—frontier labs partner to build and share resources, diverging from a zero-sum mentality and instead shoring up global competitiveness for American AI. As Gerstner observes, this détente and infrastructure kinship is vital for maintainin ...

Anthropic Achieves Unprecedented Revenue Growth in Silicon Valley

In the first four months of the year, Anthropic’s annual recurring revenue (ARR) grew from roughly $10 billion to $30 billion, tripling before soaring again to $44 billion in April. Brad Gerstner highlights that at the year’s start, Anthropic and OpenAI together were producing about $30 billion in revenue, a figure that has now reached $80 billion in four months. Forecasts suggest Anthropic could 10x this year and end 2024 at roughly $100 billion in ARR. Some discussions even project Anthropic could reach $1 trillion in ARR by 2027—a trajectory that would surpass the combined annual revenue of the current "Mag-7" tech giants (Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, Tesla), which collectively generate around $2.3 to $2.35 trillion.

David Sacks and Chamath Palihapitiya note that while giants like Google grow at approximately 20% year-over-year, Anthropic’s exponential acceleration is unprecedented, growing at rates “not 100%, certainly not 1000%.” The potential is such that if this growth continues, Anthropic could eclipse the power, valuation, and influence of the largest tech conglomerates ever assembled, transforming it from a "Mag-7" world to a "Mag-1".

This extraordinary spike in revenue is driven entirely by enterprise demand, primarily for coding and AI tools. The market is currently absorbing all of Anthropic’s expanding output, especially for coding—indicating effectively unlimited total addressable market. Brad Gerstner observes that the only thing constraining revenue is supply (not demand), namely capacity at data centers and available power. If those could scale immediately, he believes revenues would climb even more steeply.

Anthropic's Position Raises Questions About Monopoly Formation

While Anthropic and OpenAI are both posting substantial revenues, observers like David Sacks argue that Anthropic’s recent growth trajectory echoes historical monopoly formation. Sacks draws a direct analogy to John D. Rockefeller and Standard Oil, noting that companies can mask monopolistic consolidation behind a facade of public benefit and safety, and distract regulators and the public with side issues. He posits that if Rockefeller had been better at public relations—branding as “Safe Oil” and advocating for government safety regulation—the public would have missed the consolidation underway, focused instead on debates over product safety. In the same vein, Sacks suggests today's AI safety rhetoric can serve to distract from the reality of monopoly formation and regulatory capture, as companies push policies that strengthen their moats and raise barriers to entry for competitors.

Gerstner pushes back, arguing it is premature to call the current AI ecosystem a monopoly when only two companies—Anthropic and OpenAI—generate substantial revenue and are still fledgling by the standards of legacy tech giants. Yet, Sacks counters that once a company claims 80% of its market, it’s functionally a monopoly, and if Anthropic continues its exponential trend for just 18 more months, it could find itself in such an unprecedentedly powerful position.

This raises concerns that, just like Standard Oil, Anthropic (with OpenAI) could soon dominate key technology infrastructure more thoroughly than any private enterprise in history, using regulatory and policy debates to shield itself from scrutiny.

Competition Remains Theoretically Possible but Faces Structural Disadvantages

Competition in the sector is theoretically possible. OpenAI is already pivoting its focus, advancing rapidly with new models like "5.5" based on its new Spud architecture, and offering strong competition ...

White House Weighs Ai Model Vetting, Not Pre-approval

Recent reports suggest the White House is considering new oversight measures for artificial intelligence (AI) models, fueled by concerns over advanced AI such as Anthropic's Mythos model and the risks of AI-enabled cyberattacks. Jason Calacanis highlights media coverage speculating about an "FDA for AI," suggesting vetting and review for new models. However, Brad Gerstner clarifies after speaking with National Economic Council Director Kevin Hassett that the analogy to the FDA was intended to describe coordination, not an approval authority. Hassett and other officials, according to Gerstner and David Sacks, do not support a regime where every new model needs federal pre-approval before release. Instead, the media, not the administration, amplified the idea of FDA-style regulation, leading to confusion and a strong reaction in Silicon Valley.

White House Chief of Staff Susie Wiles further rejects the FDA-style approval model, reaffirming the administration’s pro-innovation philosophy. Sacks describes the president as "the most pro innovation president we've ever had" and stresses that the administration’s regulatory framework prioritizes specific legislative goals, such as cybersecurity and American competitiveness, not blanket controls or pre-release approval that could stifle innovation.

Know-Your-Customer and Api Logging Could Provide Security Guardrails

Industry leaders see know-your-customer (KYC) verification and API logging as effective, non-intrusive tools for AI safety. Jason Calacanis proposes KYC for controlling access to potent AI models, preventing malicious actors from exploiting advanced capabilities. Gerstner confirms that major AI labs already monitor API usage, track for suspicious or distillation activities, and coordinate extensively with government agencies to flag and address cyber threats. He notes the value in allowing certain uses to better understand extraction attempts and the nature of threats.

This “reasonable security” comes from robust industry self-governance and voluntary cooperation, not imposed approvals. Calacanis and Sacks emphasize self-policing among AI firms, who are motivated to set strong safeguards to avoid legal repercussions and reputational harm. The existing infrastructure relies on trust, transparency, and collaborative responses to incidents, rather than rigid government-imposed processes.

Safety Rhetoric May Justify Regulatory Capture Benefiting Incumbents

Some voices, Sacks argues, are leveraging "AI doomer" narratives and safety rhetoric to push for strict regulations that would entrench incumbent tech corporations by burdening startups with compliance overhead. He and Gerstner point out that requiring pre-release government approval would allow the state to pick AI winners and losers, slowing U.S. innovation and ceding global leadership to less-regulated competitors such as those in China. While cyber issues—such as the concerns raised by Anthropic’s Mythos—necessitate urgent system hardening in the short-term, these must not be used to permanently expand regulatory infrastructure in Washington.

Sacks warns that ideologues are using crises as opportunities to entrench regulatory capture, n ...

AI technology is fundamentally reshaping business operations, efficiency, and the broader workforce. Recent data from hyperscalers and ongoing discussions among technology investors showcase the tangible and immediate economic impact, while revealing critical questions about the sustainability and direction of AI-driven productivity gains.

Hyperscalers' Revenue Growth Reflects Enterprise Ai Demand

Major cloud providers—Amazon Web Services (AWS), Microsoft Azure, and Google Cloud—are experiencing explosive revenue growth that underscores enterprise demand for AI computing capacity. AWS grew 28% to $150 billion in revenue, Azure 39% to $108 billion, and Google Cloud an impressive 63% to $80 billion over the recent period. These companies are not only growing revenues but also expanding their operating margins, with minimal headcount growth: the MAG-5 (major tech companies) combined have only seen about 3% headcount growth over the last three years. Meanwhile, S&P 500 operating margins have improved significantly, rising from 11.8% in Q1 2024 to 13%, suggesting substantial AI-driven efficiency gains.

The sustained, month-over-month enterprise spending on AI-related coding tokens and tools is a crucial data point. As David Sacks observes, such ongoing investment only happens if enterprises are already seeing demonstrable ROI. The rapid monetization of AI tokens indicates that enterprises wouldn't be pouring in money unless there were immediate productivity benefits. This productivity is flowing through from core infrastructure, to AI models, to applications, finally reaching end users, fueling an economic boom.

Ai Delivers Cost Cuts and Productivity Gains

AI is actively cutting costs and driving measurable improvements in business performance. Companies such as Nike and DoorDash use AI to generate product imagery and photographic assets, removing the need for costly and time-consuming photo shoots. The application of AI in ad creative produces assets at half the cost while achieving double-digit percentage improvements in advertising effectiveness.

Startups especially benefit from AI coding tools, enabling them to accomplish more with fewer employees and less capital. With the aid of AI agents and automation, small teams orchestrate what used to require much larger staffs. In practice, Jason Calacanis notes that firms can now ship products at speed that would have previously required 22-person development teams, using far fewer resources. Startups gain so much value from these AI tools and tokens that they halve their hiring, stretch investments further, and deliver faster.

Critical Fork Between Divergent Ai Outcomes In 24-36 Months

The AI-driven economic boom faces a pivotal moment within the next two to three years. One possible path is that AI automation drives significant reductions in operational expenses (OpEx) and workforce size, leading to social disruption as margins increase mainly through cost-cutting. The alternate path sees OpEx stay stable (or even rise) as AI boosts productivity, revenue, and fosters new businesses and services, enhancing wider economic growth and living standards.

True transformation will require not only the upfront investment (“spending X”) but clear proof that AI adoption delivers tang ...

AI Industry Faces Negative Messaging Despite Economic Benefits

Chamath Palihapitiya observes a profound shift in public and political sentiment towards tech and particularly AI, with negativity dominating both community and political reactions. He notes that projects to increase energy supply for AI—like new datacenters—face heavy protest, with nearly half of planned capacity at risk due to public backlash, exacerbating supply constraints. Despite this, AI is driving significant economic progress: David Sacks highlights that AI is responsible for 75% of Q1 GDP growth, is fueling a construction and blue-collar boom, and is causing wage increases of 25-30% in construction.

Yet, AI’s approval remains low, and public concern does not yet rank highly among voters. Sacks cites polling showing that, despite extensive media focus on AI's risks, AI ranks only 29th out of 39 salient issues—far below concerns like healthcare and general economic welfare, which are top-of-mind for most people. The panel agrees that humans’ bias toward safety makes us hyper-aware of AI’s potential harms, from deepfakes to job loss and bioweapons.

A key factor amplifying the doom narrative is the failure of tech leaders to communicate the positive potential and realized societal benefits of AI. Palihapitiya grades tech industry leadership “D minus trending to F” in communicating AI’s upside, arguing that the lack of clear and compelling messaging on widespread social investment and uplifting American society enables negative perceptions and fear-mongering to fill the gap. Brad Gerstner and Jason Calacanis reinforce the need to better tell AI’s story and deliver clear, broad-based net benefits.

AI Wealth Fuels Public Anxiety, Sparks Regulatory Pressure

Trillion-dollar net worths accruing to a handful of founders and investors stoke public anxiety over wealth inequality and perceived power concentration. Palihapitiya underscores that the impression of a few individuals controlling the “keys” to the AI-driven future is driving backlash, with tech leaders coming across as untrustworthy or self-interested due to a lack of substantial reinvestment in society. This backdrop intensifies calls for regulation to ensure that AI-generated wealth benefits the many, rather than just consolidating among a small group at the top.

The negative sentiment, fueled by fear of job loss and socioeconomic polarization, leads to bipartisan regulatory momentum. Palihapitiya forecasts increasing oversight, whether under a Democratic or Republican administration, citing a widespread sense that AI’s economic rewards are not adequately shared with ordinary people.

Solutions: Direct Wealth Sharing and Targeted Social Investment

The panel explores several solutions to address these equity concerns and diffuse the backlash. Gerstner and Calacanis propose that AI company IPOs could voluntarily allocate 1–5% of shares to ordinary Americans—via “Invest America” accounts—so citizens directly benefit and share in AI’s compounded growth. Calacanis suggests tech leaders could pledge to donate 1% of their holdings annually for the next 20 years to causes like healthcare, education, and housing, creating a broad-based reinvestment mechanism that does not sap entrepreneurial incentive.

Raising the minimum wage is also discussed as a way for te ...