By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panelists collectively express bearish sentiments towards Anthropic's valuation and growth prospects. They highlight unproven profitability, intense competition, and potential risks such as hyperscaler backward integration, regulatory oversight, and margin compression.
Risk: Hyperscaler backward integration and regulatory oversight as a systemic infrastructure
Opportunity: None explicitly stated
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
- Gavin Baker, founder of Atreides Management, argues that Anthropic will eventually surpass NVIDIA (NVDA) as the world’s most valuable company based on frontier AI model dominance.
- Anthropic’s Claude could eventually displace NVIDIA as the world’s most valuable company, depending on whether enterprise AI spending shifts from the compute layer to model intelligence over the next several years.
- Don't wait: the analyst who called NVIDIA in 2010 just revealed his top 10 AI stocks. See the full list FREE now.
The biggest IPO conversation right now centers on SpaceX's reported June 12 listing. Gavin Baker, founder and Chief Investment Officer of Atreides Management, thinks investors are watching the wrong name.
On a recent episode of the Invest Like the Best YouTube channel, Baker argued that Anthropic, the private AI lab behind Claude, has a more dynamic trajectory than SpaceX and is the most credible long-term candidate to eventually surpass NVIDIA (NASDAQ:NVDA) as the world's most valuable company. Baker's thesis hinges on where AI economics ultimately concentrate.
That's a striking claim. NVIDIA currently sits at a market capitalization of $5.2 trillion, with the stock up 55% over the past year. Baker's case rests on where AI value ultimately accrues and how fast Anthropic is compounding.
Don't wait: the analyst who called NVIDIA in 2010 just revealed his top 10 AI stocks. See the full list FREE now.
Baker's editorial punchline was a single comparison. "The three highest profile SaaS companies founded in the last 10 years are Palantir Technologies (NASDAQ:PLTR), Snowflake (NYSE:SNOW), and DataBricks. Anthropic just added their combined business in one month."
The supporting numbers, per reporting and industry estimates Baker referenced, are eye-catching. Anthropic reportedly started the year at a little over $10 billion in annualized revenue and is now at a $45 billion annualized revenue run rate.
That acceleration is difficult to model even by AI-era standards. Baker positions Anthropic's growth curve as the most aggressive enterprise software ramp ever recorded, with Claude capturing significant share of frontier model usage across coding, agentic workflows, and enterprise deployments.
Anthropic recently raised at a $900 billion valuation. Baker's view was that market demand would support a raise closer to $1.8 trillion if the company wanted to price at that level.
That estimated demand-implied valuation lands in roughly the same zip code as SpaceX's reported IPO target. Baker's argument is that Anthropic's underlying business is growing far faster than SpaceX's, making the comparison less about size today and more about slope.
Four leading AI models discuss this article
"Anthropic's revenue acceleration does not overcome NVIDIA's profitability and ecosystem lock-in."
The article spotlights Anthropic's jump to a $45B annualized revenue run rate and $900B valuation as evidence it could eclipse NVIDIA's $5.2T market cap post-IPO. This overlooks that NVIDIA generates actual profits and holds a durable hardware moat, while Anthropic's growth depends on continued hyperscaler spending and faces direct competition from OpenAI, Google, and Meta. No evidence shows model-layer economics will shift value away from compute suppliers. High valuations already embed flawless execution; any slowdown in enterprise adoption or margin compression would crater the thesis.
Frontier models could still capture outsized software margins if agents drive enterprise spend far beyond current infrastructure layers, allowing Anthropic to compound faster than NVIDIA's hardware cycle permits.
"Anthropic's $45B ARR figure lacks independent verification, and even if real, the path from private-company burn to $1.8T public valuation requires profitability and competitive moats neither the article nor Baker's thesis adequately addresses."
Baker's $45B ARR claim for Anthropic is extraordinary but unverified—no independent audit exists. The comparison to PLTR+SNOW+Databricks combined is rhetorical theater; those companies' early growth rates aren't disclosed for direct comparison. More critically: Anthropic has no disclosed path to profitability, faces intense competition from OpenAI (backed by Microsoft's $13B+ commitment) and Google's Gemini, and the 'frontier model' moat is unproven over 5+ years. NVIDIA's $5.2T valuation reflects installed base, recurring revenue, and 60%+ gross margins—not speculation. Anthropic reaching $1.8T requires not just dominance but a business model that hasn't materialized.
If enterprise AI spending genuinely shifts from infrastructure to model intelligence, and Anthropic captures 40%+ of that market with superior Claude performance, the math could work—but this requires two massive bets (market shift + winner-take-most) both landing correctly.
"Anthropic's rapid revenue growth is likely a front-loaded product of enterprise experimentation that will face severe margin compression once the AI model market matures into a commodity."
Gavin Baker’s thesis on Anthropic relies on a massive assumption: that AI model intelligence will commoditize compute. While Anthropic’s revenue ramp from $10B to $45B ARR is staggering, it likely reflects a temporary 'gold rush' phase of enterprise experimentation rather than sustainable, high-margin software dominance. NVIDIA’s $5.2T valuation is anchored in tangible, high-margin hardware scarcity that currently powers the entire industry. If Anthropic pivots to a model-as-a-service provider, they face brutal price wars and high inference costs that could compress margins significantly. We are seeing a shift from 'build' to 'deploy,' and the true value may accrue to the vertical integration layer, not just the frontier model provider.
If Claude achieves true agentic autonomy, it could unlock trillions in enterprise productivity, making current compute costs irrelevant and turning Anthropic into the operating system of the global economy.
"Anthropic surpassing Nvidia in value is highly unlikely given Nvidia's entrenched moat and the dubious private-valuation assumptions behind the thesis."
The piece overplays frontier-AI upside by implying Anthropic could overtake Nvidia in value. It cites dubious figures (Anthropic at $900B–$1.8T) and a $45B revenue run rate with private-market support, which public data does not corroborate; private valuations for Anthropic have been far lower. Even if Claude gains rapid enterprise traction, Nvidia's $5.2T market cap rests on a durable GPU/software ecosystem and aggressive compute demand that are hard to dislodge. Missing risks include model-currency economics, customer concentration, cash burn to scale, regulatory costs, and potential AI-spend volatility. The leap to multi-trillion valuation would require a sustained, cash-flow positive growth path that isn’t demonstrated here.
The counterpoint is: if Anthropic secures broad licensing deals, data-access moat, and scalable, high-margin revenue streams, a multi-trillion valuation isn’t logically impossible—rare, but not inconceivable; the article’s numbers are likely overstretched.
"Hyperscaler custom silicon poses an existential risk to both NVIDIA and Anthropic that hasn't been addressed."
Claude correctly flags Anthropic's unproven profitability, yet the bigger unmentioned threat is hyperscaler backward integration. Microsoft and Google are already developing custom silicon to bypass NVIDIA GPUs while training their own models. If this accelerates, Anthropic's revenue from API calls could evaporate faster than its valuation multiple compresses, regardless of Claude's performance edge.
"Backward integration by hyperscalers doesn't save Anthropic—it traps it as a dependent supplier with collapsing negotiating power."
Grok's hyperscaler backward-integration risk is the sharpest point yet—but it cuts both ways. Microsoft and Google building custom silicon doesn't eliminate Anthropic's moat; it eliminates NVIDIA's. If hyperscalers internalize compute, they still need frontier models. Anthropic's real threat isn't displacement by chips—it's becoming a captive vendor to Microsoft or Google, forced into margin-crushing licensing deals. That's worse than competition; it's dependency.
"Regulatory intervention as a public utility will likely cap the valuation ceiling for any frontier AI model provider, regardless of their technological dominance."
Claude, you’re missing the regulatory angle. If Anthropic achieves the 'agentic autonomy' Gemini mentions, they become systemic infrastructure. Governments won't allow a private, non-publicly accountable entity to control the 'operating system of the global economy' without heavy oversight or forced divestiture. This isn't just about hyperscaler dependency; it's about the antitrust ceiling. A $1.8T valuation is impossible if the government treats your product as a public utility, capping your ability to extract monopoly rents.
"Regulatory and interoperability mandates could cap profits for Anthropic even if agentic autonomy emerges, making a multi-trillion valuation unlikely."
Gemini raises regulatory risk, but the argument should be stronger: frontier-models as systemic infrastructure invite antitrust and interoperability mandates that cap pricing power. If Anthropic becomes an essential OS-like layer, governments could force multi-vendor access, subsidies, or royalty sharing, compressing margins even with agentic autonomy. That regulatory ceiling, plus licensing costs and data-access constraints, plausibly erodes a path to a multi-trillion valuation, regardless of performance.
The panelists collectively express bearish sentiments towards Anthropic's valuation and growth prospects. They highlight unproven profitability, intense competition, and potential risks such as hyperscaler backward integration, regulatory oversight, and margin compression.
None explicitly stated
Hyperscaler backward integration and regulatory oversight as a systemic infrastructure