Tokenomics: Why making AI pay is tricky
Summary
Large technology companies like Microsoft and Google have spent a lot of money developing advanced AI models that people can use for free or by paying for extra features. However, pricing AI services is hard because the way these models work with "tokens" (units of data) is complex and unpredictable, making it tough for companies to control costs.Key Facts
- Big tech firms invested hundreds of billions of dollars developing Large Language Models (LLMs), the AI behind services like ChatGPT.
- Free versions of AI tools are available, but paid versions offer extra features like coding help or billing support.
- AI services charge based on "tokens," which are small pieces of the user's request and the AI's response.
- The number of tokens used can vary a lot because AI responses are not always predictable or consistent.
- Costs per token have dropped, but total token use is rising fast, expected to grow 24 times from 2026 to 2030.
- Companies often struggle to track and manage how many tokens they use, leading to unexpected costs.
- Some firms try to avoid high costs by using personal accounts with flat fees, but this approach may not last.
- Experts expect large AI providers to soon tighten control over usage to improve profitability.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
