Why this analyst says the AI bubble is 17 times bigger than the dot-com bust

A recent analysis highlights the significant market valuation of AI startups, with ten companies gaining nearly $1 trillion in market value over the past year despite lacking profitability, fueling concerns of an "AI bubble." Julien Garran of MacroStrategy Partnership contends this represents the "biggest and most dangerous bubble the world has ever seen," exceeding the dot-com and 2008 housing bubbles by multiples. Garran argues that Large Language Models (LLMs) are inherently limited, not commercially viable, and the AI ecosystem is largely loss-making, sustained only by continuous funding amidst diminishing venture capital interest. He warns that this prolonged capital misallocation will ultimately depress future GDP.

Analysis

The AI sector exhibits significant valuation concerns, with 10 unproven AI startups collectively gaining nearly $1 trillion in market value over the past year despite lacking profitability. This rapid appreciation draws comparisons to the late 1990s dot-com era, prompting MacroStrategy Partnership's Julien Garran to label it "the biggest and most dangerous bubble the world has ever seen," dwarfing prior market excesses.

Garran's analysis highlights fundamental limitations of Large Language Models (LLMs), arguing they are not commercially viable due to architectural constraints like statistical dependence, rote learning in coding, and a "scaling wall" observed since ChatGPT 4. He posits that the broader AI ecosystem, excluding Nvidia, is predominantly loss-making, sustained only by continuous funding without a clear path to profitability.

The funding landscape is shifting, with venture capital interest in highly valued AI startups reportedly diminishing. This leaves funding largely reliant on a few major players like SoftBank and sovereign entities, creating an unsustainable "permanent funding tour." The overall market sentiment is extremely negative and pessimistic, signaling a high potential for significant market impact.

This sustained misallocation of capital into non-commercially viable AI applications, if prolonged, is projected to negatively impact future GDP. While the bubble has not yet deflated and remains near all-time highs, the underlying financial fragility and diminishing investor appetite suggest an increasingly precarious market position.

Key Decisions for Investors

• Critically re-evaluate exposure to highly valued, unprofitable AI startups given fundamental commercial viability concerns and diminishing venture capital interest
• Monitor venture capital funding trends and profitability metrics across the AI ecosystem for signs of market inflection
• Consider implementing hedging strategies or reducing overweight positions in broader technology and AI-related indices to mitigate risks from potential market corrections
• Assess Nvidia's unique profitability within the context of the broader AI ecosystem's fundamental challenges and its potential dependency on continued speculative investment