Friday, May 15, 2026

Cutting Opex to Support AI Capex Probably Sets Stage for More of the Same

In a brutal way, the notion that the business value of artificial intelligence hinges in large part on displacing human labor has already begun to play out in early substitution of capital expense for operating expense.


For the top five U.S. hyperscalers (Alphabet, Amazon, Meta, Microsoft, and Oracle), combined capex is projected to approach $725  billion in 2026, a more than 60 percent increase over 2025 levels.


To fund this without destroying their balance sheets, companies are pulling two primary levers:

  1. Direct Labor Substitution: Reducing headcount in "legacy" or non-core divisions to reallocate those billions into AI infrastructure.

  2. Productivity Gains: Using AI internally ("vibe coding" or AI-assisted programming) to handle the same workload with fewer new hires.


Company 

Estimated Jobs Cut

Timeframe

Oracle

~30,000

Q1 2026

Amazon (AWS/Corp)

~16,000–30,000

Q1-Q2 2026

Meta

~16,000

Q2 2026

Dell

~11,000

Fiscal 2026

Cisco

~6,000

Q2 2026

Microsoft/LinkedIn

~8,750 (Buyouts)

Q2 2026

Google Cloud

~1,500

Ongoing 2026


Analysts at firms like Goldman Sachs and J.P. Morgan argue that while the headcount reduction is painful, the "cost of doing nothing" is higher. 


And while overspending represents a danger, firms are cutting headcount to pay for the capital investments. 


The logic is indeed "brutal" because it substitutes capex for human employment (operating expense). 


On the other hand, that is the simple logic of most computing technology deployments.


No comments:

Cutting Opex to Support AI Capex Probably Sets Stage for More of the Same

In a brutal way, the notion that the business value of artificial intelligence hinges in large part on displacing human labor has already be...