Ayham Boucher

Lecturer of information science and the executive director of Cornell’s AI Innovation Hub

Ayham Boucher, executive director of AI Strategy and Innovation at Cornell University, says the issue is not simply that AI tokens are expensive. The bigger problem is that many organizations are measuring AI adoption in the wrong way.

Boucher says:

"Over the past year, many organizations encouraged employees to use as much AI as possible. Some teams even used leaderboards or internal metrics that rewarded high token consumption. That can create the wrong incentives. Employees learn to maximize usage rather than value. The result is not responsible AI adoption, but token-maxxing.

“There is also an important distinction between different types of AI work. Using AI for many knowledge work tasks, like organizing a calendar or creating data driven insights can be far cheaper than assigning that work to a person. But software development is one of the most token-heavy uses of AI. Agentic coding tools and can consume a lot of tokens, especially when a new project is being initiated.

“The lesson for companies is not to abandon AI. The lesson is to build AI literacy. Employees need to know which model to use for which task, when a smaller or cheaper model is enough, and when to use advanced tools and models to enhance the quality of their work.

“The next phase of AI adoption will not be about who uses the most tokens. It will be about who gets the most value per token.”