Microsoft and Uber slam on the brakes of AI

2026-06-03 · Source: David Shapiro · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Fundamental Awareness, medium

Summary

Recent discussions about an "AI bubble" have intensified following reports of Microsoft canceling internal cloud usage and Uber exceeding its AI budget by an alleged \$500 million in five months due to high token costs. This has fueled claims that AI is too expensive and its market is poised to crash. However, the analysis suggests that while some AI-related assets, such as SpaceX's valuation influenced by xAI, may be overvalued, this does not signify a complete collapse. Despite approximately \$150 billion in data center build-outs being delayed or canceled this year, representing less than a quarter of the \$750 billion planned, the broader AI build-out continues. Venture capitalists and Fortune 500 CEOs remain committed to AI, viewing it as the decade's most transformative technology, indicating no major slowdown or "zeroing out" of the sector.

Key takeaway

For executives evaluating AI strategy, recognize that current cost concerns and market overvaluation do not signal an AI collapse. Your focus should remain on the long-term strategic build-out, as significant capital continues to flow into AI infrastructure and development. Do not let short-term "bubble" narratives deter your investment in this transformative technology, but rather contextualize market fluctuations against sustained corporate commitment.

Key insights

The AI market, while potentially overvalued, will not "go to zero" as the underlying build-out and strategic commitment persist.

Principles

• Asset overvaluation defines a bubble, not necessarily a crash to zero.
• Corporate and VC commitment drives AI build-out despite market fluctuations.
• Localized delays do not halt global technological transformation.

In practice

• Evaluate AI investments based on long-term build-out, not short-term market sentiment.
• Distinguish between asset valuation and fundamental technological adoption.
• Contextualize data center delays against total planned infrastructure.

Topics

• AI Market Dynamics
• AI Investment
• Cloud Computing Costs
• Data Center Infrastructure
• Technology Valuation
• Corporate AI Strategy

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Investor, Executive, Consultant

Related on AIssential

• If enough other companies report the same, the bubble pops. 🫧 — Major companies are reporting poor AI return on investment and reliability issues, suggesting a potential market correction for overvalued AI ventures.
• AI feels like this generation’s internet boom. — AI's current trajectory mirrors past tech booms, facing both skepticism and rapid industrial adoption.
• GoodData.ai CTO says the enterprise AI bubble is real, but so is AI’s transformational power — AI will become foundational infrastructure for automation, transforming work and surviving market corrections.
• The bubble is slowly popping, investment isn't able to keep up — The AI market faces a critical debate regarding its investment sustainability and the long-term viability of current pricing models.
• Is AI Doom Going Out of Style? — A nuanced "AI vibe shift" is challenging job displacement fears and re-evaluating market perceptions of AI's economic impact.
• The Stock Market Has No Idea What’s Coming — The AI market is experiencing a selloff driven by two contradictory investor fears regarding AI's impact.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by David Shapiro.