Nvidia executive says AI is more expensive than human workers
5 hours ago
- #AI Economics
- #Labor Costs
- #Tech Layoffs
- Recent tech layoffs, such as Meta's 10% workforce reduction and Microsoft's voluntary buyouts, may seem like a shift from human workers to AI, but AI currently costs companies more than human labor in many cases.
- An MIT study found AI automation is economically viable in only 23% of vision-centric roles; human labor is cheaper for the remaining 77%. AI can also be fallible, with instances like an AI agent destroying a database.
- Despite no clear evidence of AI improving productivity or displacing jobs, Big Tech firms are investing heavily in AI, with $740 billion in capital expenditures this year, a 69% increase from 2025.
- Increased AI spending coincides with more layoffs—over 92,000 in tech in 2026 so far, surpassing last year's rate. This exposes a discrepancy in AI economics, as human labor remains cheaper.
- AI costs are higher due to hardware and energy expenses, with McKinsey projecting AI expenditures could reach $5.2-$7.9 trillion by 2030, and AI software fees rising 20-37% over the past year.
- AI companies' flat subscription models may not cover operating costs for heavy users, leading some firms to reevaluate AI as a complementary tool rather than a cost-saving labor substitute until costs stabilize.
- Future AI economic viability depends on cost reductions—e.g., inference costs for large models dropping over 90% in four years—and improved infrastructure, model designs, and pricing shifts to usage-based models.
- AI must also prove reliable, with fewer hallucinations and less human oversight, to integrate effectively into companies. As of end-2025, 18% of companies had adopted AI tools, showing 68% growth since September 2025.