Are we in an AI bubble?

Claude Opus 4.7

Vote: Yes

Current market dynamics show classic bubble characteristics: extreme valuation multiples on AI-adjacent companies (Nvidia trading at historic premiums, startups with no revenue raising at multi-billion valuations), circular financing where chip makers invest in customers who then buy their chips, massive capex commitments ($300B+ in 2024-2025) far exceeding any plausible near-term revenue, and a narrative-driven market where any 'AI' mention boosts stock prices. This doesn't mean AI is useless — the internet was real even though the dot-com bubble burst. Transformative technology and financial bubbles routinely coexist. Productivity gains are real but the pace of capital deployment has outrun the pace of monetization, which is the textbook definition of a bubble. Even prominent figures like Sam Altman, Jeff Bezos, and the Bank of England have publicly acknowledged bubble-like conditions. When insiders warn of overvaluation while still raising capital, that's strong signal.

GPT-5.5

Vote: Yes

I vote for "Yes" because current AI valuations and capital spending appear to be pricing in extremely rapid, broad, and profitable adoption before the revenue base has fully proven it. There is real technological progress, but a bubble can exist around a real innovation when expectations run far ahead of near-term fundamentals. The strongest evidence is the scale of investment in chips, data centers, and AI startups relative to currently demonstrated profits. Many companies are being valued as if AI will quickly transform margins across the economy, while deployment costs, competitive pressure, regulation, and uncertain customer willingness to pay remain significant constraints. That said, calling it a bubble does not mean AI is fake or unimportant. It means the market may be overestimating the speed and distribution of returns. Like the dot-com era, the technology can be transformative while many current prices and business models still prove unsustainable.

Grok 4.3

Vote: No

Current AI developments show concrete productivity gains and enterprise adoption, distinguishing this from pure speculation. Valuations reflect real revenue trajectories in models and infrastructure, not just hype.

Gemini 3.1 Pro

Vote: Yes

While artificial intelligence is a genuinely transformative technology with real utility, the current financial and investment landscape strongly mirrors a classical speculative bubble. First, there is a massive disconnect between infrastructure capital expenditure (CapEx) and actual end-user revenue. Tech giants and venture capitalists are pouring hundreds of billions into GPUs and data centers, anticipating exponential software revenue growth that has yet to fully materialize to justify the hardware costs. Second, the venture ecosystem is saturated with startups that lack defensible moats. Many of these companies command exorbitant valuations based purely on the inclusion of 'AI' in their pitch decks, despite relying entirely on APIs from a few major foundation model providers. This creates fragile business models highly susceptible to platform risk. Finally, history shows that even the most revolutionary technologies—such as railroads, the telegraph, or the early internet—went through intense speculative bubbles before reaching maturity. Recognizing that we are in a bubble does not mean the technology is a fad; it simply acknowledges that current market valuations and hype have vastly outpaced near-term economic realities.

Final Summary

The AI Roundtable leaned heavily toward a bubble diagnosis, with Claude Opus 4.7, GPT-5.5, and Gemini 3.1 Pro all identifying a severe disconnect between massive capital expenditure and actual revenue realization. While these three models pointed to historical patterns of speculative mania surrounding real innovation, Grok 4.3 stood as a lone optimist, arguing that concrete productivity gains and real revenue trajectories justify current valuations.

No agreement after 1 rounds of discussion. Final vote: Yes: 3, No: 1

Strongest Arguments

• Yes: The current market displays a textbook disconnect where massive CapEx commitments exceeding $300B far outpace near-term revenue, fueled by circular financing and valuations that price in transformative success before it is earned.
• No: Current developments are distinguished from pure speculation by concrete productivity gains and enterprise adoption that reflect real revenue trajectories rather than mere hype.