Recent financial data reveals a widening gap in the artificial intelligence landscape; while model development labs are succeeding in reaping massive profits, companies in traditional non-tech sectors remain unable to turn these technologies into a real increase in productivity or profits.
Explosive Growth for Anthropic
The S-1 filing documents for Anthropic revealed exceptional figures reflecting the boom driven by artificial intelligence tools. The company's revenues jumped in the first half of 2026 to reach $16.3 billion, compared to just $1.2 billion in the same period of 2025. This growth is mainly due to the AI-supported coding boom launched by the Claude Code model in early 2026, which pushed major companies toward adopting models like Claude Fable 5.1 that consume billions of tokens daily.
With a valuation approaching $1.5 trillion, Anthropic's initial public offering is expected to be the largest in history. What is surprising is the massive liquidity the company holds, with more than $112 billion remaining out of the total $141 billion it has raised so far.
Infrastructure and Compute Challenges
Despite this financial abundance, Anthropic faces major operational challenges; the company is committed to spending hundreds of billions of dollars on cloud infrastructure with giants like Google, Amazon, Microsoft, and SpaceX. The company is expected to face compute-constrained limitations over the next year. To cut costs, the company may be forced to look for new, less reliable providers, while relying more heavily on natural gas-powered data centers.
The Productivity Gap in the General Economy
In contrast, economic experts offer a different vision of the technology's trajectory. Torsten Slok, chief economist at Apollo, points out that artificial intelligence has not changed the rules of the game outside the tech sector. While the "Magnificent 7" companies are driving the US economy forward, other major companies (referred to as the S&P 493) show no signs of rising profits or achieving tangible productivity gains resulting from the use of artificial intelligence.
Data reveals that 80% of the top 1% of AI spenders in the United States are venture capital-backed (VC-backed) companies, reinforcing the idea that the benefits of this technology are still confined to the halls of Silicon Valley and have not yet reached traditional companies in the public market.
What is Changing Practically?
This contrast presents us with a new technical and economic reality; investors and companies face a fundamental question: Are we only in the infrastructure-building phase, or will the return on investment (ROI) soon reach non-tech sectors?
- Spending Concentration: Massive investments are concentrated in model and software development, while its application in traditional industries remains of limited impact so far.
- Dependency: Startups and labs consume massive computing resources, putting pressure on data center and energy supply chains.
- Investment Risks: The gap between AI lab profits and the lack of productivity in other companies raises questions about the sustainability of this growth if it does not translate into comprehensive economic value.