Tokenomics: AI Has an income statement - it’s time to read it

The hyperscalers, such as Amazon, Google, Meta, and Microsoft, are investing over $500 billion in capital expenditure for AI infrastructure this year. This indicates a strong demand for AI capabilities, which will likely lead to recouping their investment and more. The revenue from AI applications is expected to be a major source of cash flow for companies in various sectors, including industrial ones like manufacturing, energy, and transportation.

While much discussion around AI costs has focused on training large language models, the ongoing inference tasks, which apply models in live environments, are a crucial aspect that is often overlooked. Inferences are continuous and can significantly increase a company's IT budget as they scale AI deployment across their operations.

A new economic approach, "tokenomics," is needed to accurately measure and charge for AI consumption. Tokens serve as the units through which generative AI systems process and generate information, similar to how kilowatt-hours are charged to electricity customers. As companies increasingly adopt agentic systems, which require numerous model calls and extensive processing, these costs will rapidly increase.

The AI inference market is projected to more than double in the next five years, mirroring a similar growth trend during the transition to cloud computing in the 2010s. While cloud computing led to sharply decreasing unit costs and skyrocketing total usage, resulting in higher overall IT spending, AI tokenomics is accelerating at a faster pace.

CFOs are familiar with cloud computing economics and can optimize variable cost structures. However, the sheer volume of tokens generated by AI in industrial environments will far exceed any efficiency gains, leading to substantial total costs. Lighter-weight models, edge-based processing, and other techniques can help mitigate these costs. Overall, while AI holds significant upside potential for industry, the associated costs should be carefully considered and accounted for in financial planning.