Why the AI Boom May End in 2026: What It Means for Markets

David Woo Unbound

7,579 views • yesterday

Video Summary

The US economy's resilience in 2025 was significantly propped up by an AI bubble, which fueled 92% of GDP growth and supported household spending. While many on Wall Street anticipate a continuation of this trend with strong equities and a weaker dollar in 2026, this video argues that the AI capex boom is far from certain. OpenAI's massive $1.4 trillion compute infrastructure commitment, initially seen as a positive indicator, is now questioned following Google's Gemini 3.0 release, which has outperformed OpenAI's offerings and shifted momentum. The emergence of highly capable, low-cost AI models from China further challenges the established players and their inflated valuations. One highly interesting fact is that Chinese AI models like Kimmy K2 and Deepseek 3.2, which rival or surpass leading models in benchmarks, cost a fraction of what OpenAI plans to invest.

Short Highlights

The US economy's GDP growth in the first half of 2025 was 92% driven by AI-related capex spending.
OpenAI has committed to investing $1.4 trillion in compute infrastructure over the next eight years.
Google's Gemini 3.0 has been released and reportedly outperforms ChatGPT in several key areas.
Chinese AI models like Kimmy K2 and Deepseek 3.2 are emerging as strong, low-cost competitors.
The video suggests that in 2026, the market will shift focus from AI capex spending to how AI companies will monetize their technology, which is predicted to be more difficult.

Key Details

The AI Bubble's Role in the US Economy [00:00]

AI-related capex spending contributed to 92% of US GDP growth in the first half of 2025.
The wealth generated by the AI rally supported household spending, preventing a potential recession in 2025.
If AI was crucial in 2025, its importance is expected to remain in 2026.

"it is not unreasonable to say that the US economy would have gone into a recession in 2025 if it hadn't been for AI."

Questioning the Continuation of the AI Capex Boom [01:16]

Wall Street forecasts for 2026 assume the AI boom will continue and productivity gains will lower inflation.
The video expresses skepticism about the continued AI capex boom into 2026, viewing the bursting of the AI bubble as a central scenario rather than just a risk.
The AI bubble has been fueled by soaring AI-related capex spending, with investors equating this spending with future returns, leading to correlated stock prices and valuations.

"I think this is much less certain than it looks."

OpenAI's Ambitious Commitments and Market Reaction [03:09]

OpenAI has announced significant partnerships with chip manufacturers (Nvidia, AMD, Broadcom, Foxconn) and hyperscalers (Amazon, Google, CoreWeave, Oracle).
These deals involve a commitment of $1.4 trillion in compute infrastructure over the next eight years.
Despite OpenAI's large revenue ($12 billion) and these massive commitments, the market initially reacted positively, equating spending with future returns and overlooking the financial viability.

"Once again, the market decided to equate spending with future returns."

The Impact of Gemini 3.0 and Intensifying Competition [04:09]

Google's Gemini 3.0 release has received universally positive reviews, outperforming ChatGPT in areas like multimodal reasoning and complex instruction following.
This development shifts momentum to Google, potentially making OpenAI's trillion-dollar capex promise seem overblown if they are no longer the undisputed leader.
Competition is also emerging from China, with startups like Moonshot releasing models like Kimmy K2 for only $4.6 million, which outperformed ChatGPT 5.0 on benchmarks.
Deepseek has released free models (Deepseek 3.2 and 3.2 Special) that closely match frontier models in reasoning and coding.

"The problem for the AI bubble is that if OpenAI is no longer the undisputed leader, then it trillion dollar capex promise looks like a lot of hot air."

Monetization Challenges and Valuation Concerns [05:47]

The market is underestimating the difficulty companies like OpenAI face in monetizing their AI technology.
As competition intensifies, prices are expected to fall, leading to decreased expected returns.
This will force a re-evaluation of AI-related capex spending by the market.

"My thinking has not changed. As competition intensifies, prices will fall and so will the expected returns."

Implications for Microsoft, Tesla, and Nvidia [06:07]

Gemini 3.0 poses a threat to Microsoft, whose AI products rely on OpenAI's foundational models. Microsoft's efforts to develop its own LLMs (MAI) may not be enough to close the gap with Google.
Tesla, which markets itself as an AI and robotics company, faces a significant problem if Gemini 3.0's advanced multimodal reasoning models outperform its own FSD stack, potentially giving Google a lead in robo taxis.
For Nvidia, the threat lies in Google training Gemini on its own TPUs, meaning Nvidia won't benefit from Google scaling its data centers. Furthermore, Chinese AI models demonstrate that frontier models can be trained without expensive Nvidia chips, as the scaling is algorithm-bound.

"Well, Microsoft tried to compensate by increasing his AI related capex spending further. I cannot rule it out but the market will see through it very quickly..."

The 2026 Outlook: From Capex to Monetization [08:24]

The AI trade in 2025 was centered on capex spending.
In 2026, the market is expected to shift focus to how AI companies will generate revenue, which is anticipated to be more challenging than spending money.
The current valuations of the AI sector are considered difficult to justify.
When the market realizes this, it will have a significant impact on US stocks, interest rates, and the US dollar, with the bursting of the AI bubble being bullish for steepeners and bearish for the dollar.