Perplexity's Comet promised to tame email chaos. It drafted bland responses to critical messages and returned dead LinkedIn links. Not alone—AI agents keep failing basic tasks while investors cheer. Tests reveal systematic gaps between pitch and reality.
Good Morning from San Francisco, Enterprise AI hit profitability this week. Companies celebrate returns while cutting the training budgets that
Enterprises report 74% positive AI returns while cutting training budgets 8%. The Wharton study reveals companies extracting productivity gains today by depleting tomorrow's capabilities—a business model that works until skills erode.
Explore the implications of Microsoft's $80 billion investment in AI infrastructure. Is a strategic shift on the horizon? Stay informed here.
Microsoft is quietly backing away from its ambitious AI data center expansion. The tech giant has canceled leases totaling hundreds of megawatts with multiple private data center operators across the United States, according to TD Cowen analysts.
This strategic retreat includes halting lease conversions, abandoning 100-megawatt capacity deals, and letting over a gigawatt of larger site agreements expire. Microsoft has also redirected significant international spending back to the U.S.
The pullback seems directly connected to OpenAI, in which Microsoft holds a major stake. Analysts believe Microsoft overestimated infrastructure needs related to OpenAI's growth. A Wisconsin data center construction pause, initially meant for OpenAI workloads, supports this theory.
What makes this situation perplexing is Microsoft's recent announcement to invest over $80 billion in AI-enabled data centers this fiscal year. CEO Satya Nadella had emphasized the need to "sustain spending to meet exponentially more demand" during a January earnings call.
When questioned, Microsoft confirmed its fiscal year spending target but avoided addressing the specific lease cancellations. A spokesperson stated, "While we may strategically pace or adjust our infrastructure in some areas, we will continue to grow strongly in all regions."
This apparent contradiction hasn't gone unnoticed. Nadella recently suggested that AI supply meeting demand would be evident if global economic growth accelerates. His comments frame the issue within broader economic conditions.
The ripple effects are already visible. European energy stocks took substantial hits, with Schneider Electric falling 7.2% and Siemens Energy dropping over 10%. Meanwhile, Microsoft and Nvidia shares rose in premarket trading Monday.
Investor skepticism has intensified following breakthroughs by Chinese AI startup DeepSeek. They've showcased technology matching or exceeding Western rivals at a fraction of the cost. This development questions the economic justification for extensive data center investments.
Critics have long pointed to the scarcity of practical AI applications despite billions in infrastructure investments. The gap between investment scale and marketable applications grows more glaring daily.
Microsoft's changing strategy might also reflect evolving partnerships. OpenAI and SoftBank announced a joint venture to invest up to $500 billion in data centers and AI infrastructure in January. This arrangement could reduce Microsoft's need for its own extensive infrastructure.
GitHub isn't building the best AI coding agent—it's building the layer beneath all of them. Agent HQ bets enterprises value coordination over specialization. Platform strategies that extract value from ecosystem competition reshape markets more durably.
OpenAI commits $1.4T to infrastructure; Nvidia projects $500B in chip sales. Same day, same pitch: AI requires sovereign-scale capital. The tension: revenue is still catching up to the rhetoric, and someone will hold expensive capacity if demand falters.
OpenAI's restructure puts a $130B nonprofit in control of a $500B for-profit. Microsoft locked in 27% and access through 2032. But a year of regulatory review produced new numbers, not new transparency—and Microsoft's accounting gaps remain unaddressed.
Anthropic's $7B revenue run rate trails OpenAI's $13B, but the gap narrows when accounting for customer mix. New Excel integration and financial connectors target Wall Street as Microsoft embeds both AI providers—hedging infrastructure bets in an uncertain market.
Get the 5-minute Silicon Valley AI briefing, every weekday morning — free.
