tech
Perspective: AI demand is inflated, and only Anthropic is being realistic
The main usage metric for artificial intelligence, called tokens, looks explosive on paper, but it may be significantly overstated.
TL;DR
- Token consumption, the primary metric for AI usage, may be an inflated indicator of actual demand.
- Companies are incentivizing employees to use more tokens rather than focusing on productivity, leading to distorted usage patterns.
- Anthropic is shifting to per-token billing to align revenue with actual usage, preparing for a potential AI demand correction.
- Flat-rate pricing models, common in early AI adoption, are becoming unsustainable with increased agentic AI usage.
- Anthropic has cut off third-party tools that were excessive token consumers and is changing its enterprise contract models.
- Other companies, like Salesforce, are exploring alternative metrics beyond token count, such as 'agentic work units'.
- Anthropic's pricing strategy provides cleaner data for potential investors, while competitors may face challenges proving the real value of their numbers.
- The market's perception of AI demand will be crucial for companies pursuing IPOs.