Corporate America’s enthusiasm for artificial intelligence (AI) is facing challenges as companies grapple with the complexities and costs associated with purchasing and using “tokens,” the units of computing power underlying AI services. Initially, organizations embraced aggressive AI adoption—known internally as “token-maxxing”—to take advantage of its promised efficiencies and innovations. However, the swift depletion of budgets, a phase described as “tokenmining,” has led to increased scrutiny and strategic reassessment at both engineering and executive levels.

Tokens function as metered units of AI usage, often equated to a few characters of text processed by AI models. They serve as a form of currency for buying access to computational power, which is generated from electricity and semiconductor resources. Despite the centrality of tokens to AI consumption, businesses face significant uncertainty in understanding their actual value, pricing structures, and the return on investment they deliver.

Experts say this environment is giving rise to a new area of study, dubbed "tokenomics," which focuses on how tokens are produced, traded, and converted into economic value. Howard Rubin, an economist advising firms on technology spending, noted that companies currently treat AI expenditures as investments but face considerable risk due to a lack of established accounting practices and uncertainty about outcomes.

Comparisons have been made between tokens and traditional commodities like petroleum products, where crude oil is refined into fuels measured by standard units such as British thermal units and traded transparently on wholesale markets. However, this analogy only extends so far. Unlike fuel commodities, the AI ecosystem involves thousands of models with varying capabilities, little transparency regarding usage and pricing, and no consensus on metrics such as the energy cost per token or the optimal number of tokens required for specific AI tasks.

The rapidly evolving AI landscape presents unique challenges for organizations attempting to balance potential benefits with financial sustainability. As companies continue to explore how best to integrate AI technologies, understanding token consumption and establishing clearer valuation frameworks remain critical priorities.