Authors: Scott Galloway / Ed Elson / Mia Silverio
Translation: TechFlow
TechFlow Introduction: Nearly 50,000 people have been laid off this year due to AI, but more and more companies are finding that the cost of using AI is higher than human labor. Uber burned through its entire 2026 AI budget in four months, Microsoft is cutting Claude Code licenses in multiple departments, and an Anthropic employee used up $150,000 in API credits in a single month. Scott Galloway believes companies will ultimately turn to Chinese large language models that are 10-30 times cheaper, which will force Trump to impose restrictions.
Is AI More Expensive Than the People It Replaces?
Nearly 50,000 employees have been laid off this year, citing AI as the reason. This figure nearly equals the total for all of 2025. For companies adopting AI, the logic is simple: AI can do the jobs people do.
But in recent weeks, this logic has hit a wall. More and more companies are discovering that the actual cost of using AI is higher than the human labor it is intended to replace.
Figure: The AI Cost Shock for Businesses – AI spending and cost feedback from companies like Uber, Microsoft, Nvidia, Meta
Uber burned through its entire 2026 AI budget in just four months. The COO said it's becoming increasingly difficult to justify AI expenditures internally. Microsoft is cutting Claude Code licenses in multiple departments for one simple reason: cost.
A Nvidia executive stated that compute costs now "far exceed employee costs." Meta, Pinterest, and Spotify all cited rising inference costs as a drag on margins in their Q1 earnings reports.
How big are corporate AI budgets? A survey by cloud cost management company CloudZero shows that in 2025, 45% of businesses spent over $100,000 per month on AI, up from only 20% the previous year.
An even more extreme case within Anthropic: one employee spent $150,000 on Claude Code in a single month. For that to be financially justifiable, this engineer would need to do the work of 11 average engineers.
In the current market, the performative value of the word "efficiency" has been consistently rewarded, to the point where companies don't even need to actually calculate efficiency. 79% of S&P 500 companies mentioned AI in their recent earnings calls, but only 8% disclosed any AI-related revenue.
Figure: S&P 500 Companies' AI Rhetoric vs. Actual Revenue Disclosure
The same CloudZero report also found that only half
Chinese Large Models Will Be the Biggest Winners
Scott Galloway's judgment is: companies will ultimately turn to the cheapest models, which are Chinese large language models. Chinese models are 10 to 30 times cheaper than American models.
Data is already validating this trend: the share of Chinese models in developer usage surged from about 1% in 2024 to over 60% in May of this year, and 80% of US AI startups are using Chinese open-source AI models.
Figure: Changes in Share of Chinese Large Models in Developer Usage & Usage by US AI Startups
