Khosla's AI Affordability Thesis: The $10k-$30k Buying Power Prediction vs. Current Data

The core promise of Vinod Khosla's thesis is a "hugely deflationary" shift. He predicts that by 2040, $10,000 to $30,000 could buy far more than $100,000 does today, as AI makes labor costs near zero. This vision hinges on the idea that automation will drive such dramatic economic abundance that the real value of money expands exponentially. OpenAI CEO Sam Altman frames this similarly, arguing AI-driven productivity will introduce strong deflationary forces that lower the cost of goods and services.

This prediction assumes that the benefits of AI productivity are widely shared. However, a major risk is outlined by Citi, which warns that uneven incomes could cause prices to fall as some pull back on spending. The bank sees a precarious path where AI job losses pile up, potentially leading to deflation. This scenario is distinct from the idealized abundance Khosla envisions, as it hinges on a "small AI elite" capturing most gains, leaving broader demand weak.

The timeline for this shift is aggressive. Khosla believes most labor will become automated within 15 years, with AI capable of doing 80% of jobs by the early 2030s. The transition from today's $100k purchasing power to tomorrow's $10k-$30k equivalent would be rapid and disruptive, fundamentally altering the relationship between income, spending, and price levels.

The Data Disconnect: No Deflation, Just Mixed Price Pressures

The current economic flows show no sign of the deflationary boom Khosla predicts. Consumer inflation cooled to 2.4% in January, while wholesale prices rose faster at 2.9%. This disconnect suggests broad-based price declines are not underway. Instead, the data reveals a more complex picture of growth without labor market expansion.

Strong GDP growth has come alongside negligible job creation. In 2025, the economy added an average of just 15,000 new jobs per month, barely above zero. This gap between output and employment is often cited as a potential sign of a productivity surge. Yet the productivity data itself is noisy and shows no unusual jump. Over several quarters, productivity growth has averaged about 2.2%, which is strong but not dramatically different from pre-pandemic trends.

The bottom line is that today's flows do not confirm a massive AI-driven productivity boom. The weak job growth could stem from other factors like compositional shifts in the workforce or measurement quirks. Until productivity data shows a clear, sustained acceleration, the evidence for a near-term deflationary shift remains speculative. The current setup is one of mixed pressures, not the wholesale price collapse implied by the affordability thesis.

The Capital Flow Reality: AI Investment vs. Economic Impact

The money flowing into AI infrastructure is staggering, but its real-world economic effect is proving elusive. U.S. data-center spending alone is expected to exceed half a trillion dollars in 2025. This massive build-out, driven by hyperscalers, has seen actual capital expenditure consistently outpace analyst estimates. For two years running, spending has exceeded 50% growth, with consensus projections for 2026 now at $527 billion.