Businesses using AI services from the mentioned companies face very high expenses. The practice of "tokenmaxxing," which implies making use of as many AI tokens as possible, has become unpopular now.
Meta, Uber, Walmart, Amazon, and other businesses are trying to control the AI usage since their expenses associated with AI applications are increasing rapidly. In particular, Uber exhausted its entire annual budget for AI use within only four months.
Now, companies are moving towards "tokenminning" (token minimizing) or efficient use of AI.
According to experts, many businesses promoted the token usage by their employees since they had no way of assessing AI productivity. Such approach led to higher expenses but not necessarily to better results.
AI service providers charge companies depending on the use of their services and number of tokens consumed. Simple operations consume a few tokens while complex coding processes can consume thousands or even millions of tokens.
With time, more sophisticated AI models become more expensive. Employees tend to choose the most advanced AI models for any task they face.The rise of AI agents, which can complete tasks for hours without human input, has increased usage even more. Now companies are focusing on smarter and more cost-effective ways to use AI.
