INSERT COIN

Enjoying this bite?

Sign in (free) to track this channel, unlock new bites the moment they drop, and search every summary we've ever made.

See Channel

Is the AI Bubble About to Be Tested?

Published 2026.09.26

Watch on YouTube Channel on TubeBite

0:00 / 0:00

Source: YouTube. Summary is AI-generated from the video's captions and may contain errors. It does not represent the views of TubeBite, the creator, or YouTube. Watch the original before relying on anything important.

SUMMARY

Patrick Boyle, hedge fund manager and finance professor, analyzes the looming IPO of AI company Anthropic and the broader implications for the artificial intelligence sector's valuations. He explores the challenges of justifying sky-high valuations, the circular financing among industry giants, and the risks posed by rising interest rates and intensifying competition.

MAIN POINTS

Anthropic is reportedly seeking a $2 trillion IPO valuation amid a record-high NASDAQ and strong US business growth, but IPOs are being delayed or canceled.
Historical context from the dotcom bubble highlights the dangers of paying high revenue multiples, with Anthropic's proposed valuation far exceeding even those extremes.
Speculative forecasts and inflated total addressable market (TAM) estimates are used to justify AI company valuations, with some predictions reaching tens of trillions of dollars.
Circular financing structures emerge as companies like SoftBank, OpenAI, and Nvidia engage in complex arrangements to support each other's valuations and operations.
Despite Nvidia's dominant position and soaring profits, its stock trades at historically low multiples, reflecting market skepticism about the sustainability of current AI spending.
A price war among AI labs, falling model costs, and the rise of cheaper alternatives challenge the ability of leading companies to maintain premium pricing and high margins.
The upcoming Anthropic IPO will provide the first public test of pure AI lab valuations, with significant implications for major investors like SoftBank, Amazon, and Google.

DETAILED ANALYSIS

Anthropic, a leading artificial intelligence company, is preparing for what could be the largest initial public offering (IPO) in history, targeting a valuation of approximately $2 trillion. This figure rivals the combined value of the ten largest tech IPOs to date and comes at a time when the NASDAQ index is at record highs and US business activity is surging. However, despite favorable market conditions, several high-profile IPOs have been delayed or withdrawn, reflecting underlying investor caution.

Notably, Anthropic's public filing, initially expected in late August, has yet to materialize, and OpenAI has postponed its own listing to the following year.

The valuation process for Anthropic is complicated by the lack of comparable public companies and the difficulty of forecasting cash flows in a rapidly evolving sector. Traditional methods such as discounted cash flow analysis are challenged by the company's explosive but volatile revenue growth and the current high-interest-rate environment. The yield on the 10-year US Treasury recently reached its highest level since 2004, increasing the discount rate applied to future profits and raising the cost of capital for companies building out expensive infrastructure like data centers.

This has contributed to the postponement of several tech and infrastructure IPOs, including those in the energy and data center sectors.

Historical parallels are drawn to the dotcom bubble, particularly through the lens of Scott McNealy's famous critique of excessive revenue multiples during the Sun Microsystems era. At the peak of the dotcom boom, investors paid up to ten times revenue for tech stocks, a level McNealy argued was unsustainable even under unrealistically optimistic assumptions. Anthropic's proposed valuation implies a multiple of around 31 times revenue, far exceeding those historical benchmarks.

Even with highly favorable assumptions—no costs, taxes, or capital expenditures—the present value of all future revenues falls significantly short of the $2 trillion target, especially as higher interest rates further reduce the value of distant profits.

To bridge this gap, proponents of high AI valuations increasingly rely on the concept of total addressable market (TAM). Recent filings and analyst reports have cited TAM figures for AI and related sectors ranging from $22 trillion to $60 trillion, often representing a substantial portion of global economic output. These numbers have grown rapidly, sometimes outpacing both company revenues and the broader economy.

However, history shows that companies rarely capture more than a small fraction of their TAM, as illustrated by Uber and WeWork, whose lofty market projections failed to translate into sustained profitability or market dominance. Some AI industry forecasts even venture into speculative territory, suggesting scenarios where AI-driven economic growth renders traditional financial metrics obsolete.

Despite the optimism, there are significant practical and financial challenges. Companies like SB Energy, SoftBank's US data center developer, have sought valuations as high as $50 billion without having operational facilities. Their business models depend on massive capital expenditures and complex financing arrangements, including high-yield (junk) bond offerings and deferred dividend payments.

These projects often face construction delays, regulatory hurdles, and the risk that technological obsolescence will erode asset values before they can recoup their investments. The chips that power AI data centers depreciate rapidly, making long-term infrastructure financing risky.

A notable feature of the current AI investment landscape is the circular nature of financing among major players. SoftBank, OpenAI, and Nvidia are deeply intertwined through equity stakes, debt guarantees, and long-term contracts. For example, SoftBank raises capital at high interest rates to fund OpenAI, which in turn leases data center capacity from SB Energy, providing revenue to justify SB Energy's valuation.

Nvidia invests in SB Energy and guarantees financing for its projects, ensuring its hardware is used exclusively. This interdependence creates a system where valuations and revenues are mutually reinforcing, but also potentially fragile if any link in the chain falters.

Amid this, Nvidia stands out as the primary supplier of the hardware underpinning the AI boom. Its sales have surged from $27 billion to an estimated $410 billion in four years, and net income is expected to nearly double. Yet, Nvidia's stock trades at less than 17 times forward earnings, its lowest valuation in over a decade.

Several factors contribute to this paradox: the market views Nvidia as a cyclical company at a potential earnings peak; its customers are building competing hardware; and much of its revenue is tied to cash-burning AI labs whose long-term viability is uncertain. Additionally, the ability to short Nvidia stock and the transparency of its public market valuation contrast sharply with the opaque, optimism-driven pricing of private AI labs like Anthropic.

The competitive landscape is intensifying, with price wars erupting among AI labs. Companies such as Anthropic and OpenAI have slashed prices for their latest models by up to 50%, while new entrants like Typesafe AI offer specialized models at a fraction of the cost. The cost of achieving a given level of AI performance has dropped dramatically, outpacing even other transformative technologies.

However, while the costs of inputs like chips and electricity are rising, the prices that AI labs can charge for their outputs are falling, compressing margins and challenging the sustainability of current business models.

Despite these headwinds, usage of AI tools continues to grow rapidly, and some analysts believe that companies able to integrate models with proprietary software may achieve durable profitability. Nevertheless, the upcoming Anthropic IPO represents a critical test for the sector. It will establish a public market valuation for a pure AI lab for the first time, with significant consequences for major investors such as SoftBank, Amazon, and Google, whose balance sheets are tied to private valuations.

The outcome will also influence broader market sentiment, as academic research suggests that periods of heavy share issuance often precede underperformance for both new issuers and the overall market. Ultimately, the sustainability of the AI boom will depend on whether these companies can deliver profits that justify their ambitious valuations, or whether investors will look back and question the assumptions that drove the current frenzy.

LINKS

KEYWORDS