While the cost-per-token is decreasing as models become more efficient, this efficiency gain drives a massive increase in new use cases and overall consumption. This economic principle, Jevons Paradox, explains why total enterprise spending on model inference is skyrocketing, even as the unit cost falls.

The massive AI spending from hyperscalers and enterprises isn't justified by current profits or clear ROI. Instead, it's a defensive, game-theoretic move driven by the fear of being technologically outmaneuvered if competitors achieve a breakthrough first.

When a non-tech firm like Oreo's parent invests a disproportionately large amount of its budget ($40M) on a proprietary AI model, it may indicate a vanity project. This spending is often driven by executives seeking to appear innovative rather than by a sound business case.

In the current 'capability exploration' phase, companies incentivize developers to use as many AI tokens as possible. This serves as a visible, albeit inefficient, signal of AI adoption to management, prioritizing quantity over quality.

The AI market has cleared its first ROI hurdle: model revenue has justified massive infrastructure investment. Now it faces a second, harder test. Enterprises spending billions on AI tokens must demonstrate tangible financial benefits, like higher margins or revenue, to sustain the flywheel.

The narrative of insatiable AI compute demand is partially a bubble. It's fueled by inefficient early models ("token maxing") and a culture where tech executives brag about their AI spending as a status symbol, a behavior not seen with traditional cloud costs. This suggests demand could normalize.

While the cost for GPT-4 level intelligence has dropped over 100x, total enterprise AI spend is rising. This is driven by multipliers: using larger frontier models for harder tasks, reasoning-heavy workflows that consume more tokens, and complex, multi-turn agentic systems.

While most of the AI market will gravitate towards cheap, 'good enough' open-source models, Anthropic is capturing a lucrative high-end segment. These users are willing to pay significantly more for even marginal improvements in performance, creating a durable 'luxury token' niche.

Current AI models suffer from negative unit economics, where costs rise with usage. To justify immense spending despite this, builders pivot from business ROI to "faith-based" arguments about AGI, framing it as an invaluable call option on the future.