Since the launch of ChatGPT in 2022, AI rapidly replaced human labor across industries due to its low cost and high efficiency, driving mass adoption and layoffs. However, widespread enterprise usage, token-based pricing, and practices like “tokenmaxxing” have inflated demand and costs.
Supply constraints in data centers and rising compute prices have further increased token costs, more than doubling since 2025. Companies like Meta and Microsoft have reported massive usage and budget strain.
As AI expenses approach or exceed labor costs in some roles, enterprises are increasingly questioning return on investment and the long-term economic viability of large-scale AI deployment.