Authors: Scott Galloway / Ed Elson / Mia Silverio

Translated by: Deep Tide TechFlow

Deep Tide 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 labor. Uber burned through its annual AI budget in four months, Microsoft cut Claude Code licenses in multiple departments, and an employee at Anthropic used $150,000 worth of API credits in a month. Scott Galloway believes that companies will eventually turn to the 10-30 times cheaper Chinese large models, which will force Trump to take action to impose restrictions.

Is AI more expensive than the people it replaces?

Nearly 50,000 employees have already been laid off this year on the grounds of AI. This number is almost at the same level as the total for the entire year of 2025. The logic for companies adopting AI is simple: AI can do the work that 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 labor it is meant to replace.

Image: Impact of AI costs on companies — AI spending and cost feedback from companies like Uber, Microsoft, Nvidia, Meta, and others

Uber burned through its entire annual AI budget for 2026 in just four months. The COO said it has become increasingly difficult to justify AI expenditures internally. Microsoft is cutting Claude Code licenses in several departments for one reason: expensive.

Nvidia executives said that the cost of computing power now "far exceeds employee costs." Meta, Pinterest, and Spotify listed rising inference costs as a drag on profitability in their first-quarter reports.

How large is corporate AI budgeting? A survey by the cloud cost management company CloudZero shows that by 2025, 45% of companies will spend over $100,000 on AI each month, up from only 20% the previous year.

There is an even more exaggerated case at Anthropic: one employee spent $150,000 on Claude Code in one month. To make this expense sustainable, this engineer would need to accomplish the work of 11 typical engineers.

In the current market, the word "efficiency" has been so rewarded for its performance value that companies do not even need to calculate efficiency genuinely. In the S&P 500, 79% of companies mentioned AI in their recent earnings calls, but only 8% disclosed any AI-related revenue.

Image: Comparison of AI discussions and actual revenue disclosures among S&P 500 companies

The same CloudZero report also found that only half of the surveyed companies said they could confidently assess the returns on their AI investments. Match Group CEO Spencer Rascoff stated that AI costs the company $5 million to $10 million a year. When asked about ROI, he said, "I feel we benefit from it, but it's hard to feel."

The Chinese large model will be the biggest winner

Scott Galloway's assessment is: companies will eventually turn to the cheapest model, which is the Chinese large model. The pricing of Chinese models is 10 to 30 times cheaper than that of American models.

Data is already validating this trend: the share of Chinese models in developer usage soared from about 1% in 2024 to over 60% in May of this year, with 80% of American AI startups using Chinese open-source AI models.

Image: Changes in the share of Chinese large models in developer usage and the usage situation of American AI startups