The idea that the AI industry is in a speculative bubble has become conventional wisdom in online discussions, with many assuming a collapse is imminent. However, according to a developer’s analysis, this assumption may not be reasonable—there is a realistic chance that AI companies’ rapid growth is justified and they have room for further expansion.
A key point in the anti-bubble argument concerns how AI companies spend money. While the industry has committed to spending at least 3 trillion dollars and has already spent an estimated 1.9 trillion, these expenses are primarily one-off infrastructure purchases rather than recurring operating costs. According to the analysis, once data centers are built, AI spending will plummet significantly. Outside of infrastructure buildout, AI is highly profitable: Anthropic’s regulatory filing shows the company makes $2.08 for every $1 of compute spent. The company has also begun earning more than its combined compute and infrastructure investments, suggesting the data center construction model is sustainable.
The debt taken on to fund infrastructure represents a serious long-term obligation. If the industry must pay off 5 trillion dollars over 20 years, that equals 250 billion annually—a substantial sum but comparable to what major tech companies already generate. Anthropic recorded 72 billion dollars in losses via share conversions in the first six months of 2026, but this cost reflects the company’s skyrocketing 1 trillion dollar valuation.
Revenue growth across the industry has been extraordinary. Anthropic’s revenue grew from 41 million in 2023 to 16.3 billion in the first six months of 2026, suggesting similar trajectories at competitors like OpenAI. The analysis argues this growth can continue, pointing to massive unmet demand: programmers and mathematicians lag behind AI systems, lawyers and doctors could benefit from AI-assisted paperwork, and self-driving cars are becoming viable. Even without further AI improvements, adoption could expand consumption tenfold.
Regarding competitive dynamics, while no single AI company can maintain lasting dominance—as a leaked 2023 Google memo acknowledged—this does not preclude profitability. The analysis compares this to the chicken sandwich wars among fast-food chains, where intense competition and lack of durable advantages did not prevent strong profits. Additionally, most users will not deploy locally hosted models, meaning only a few mainstream options will capture the market, enabling brand loyalty and customer retention.
Key facts
- AI companies have committed to spending at least 3 trillion dollars, with an estimated 1.9 trillion already spent
- Anthropic’s S-1 filing shows the company makes $2.08 for every $1 of compute, indicating strong profitability outside infrastructure costs
- Anthropic’s revenue grew from 41 million in 2023 to 16.3 billion in the first half of 2026
- If the AI industry must repay 5 trillion dollars over 20 years, that equals 250 billion annually—comparable to major tech company revenues
- The analysis argues massive unmet demand exists for AI applications in programming, mathematics, law, medicine, and autonomous vehicles