Recent reporting highlights a paradox in the AI market. TechCrunch's analysis of consumer AI economics shows that while products like Meta's Muse and OpenAI's Dots attract attention, the percentage of consumers paying for AI services is tiny—around 2.2% as of May, with average monthly spend of $31. Growth in both adoption and spending is linear, barely moving even as models improve dramatically. Meanwhile, Tom's Hardware cites a report suggesting the price of AI "intelligence" is crashing faster than Moore's Law, outpacing cost declines seen in compute, DNA sequencing, and lithium batteries.
These two findings are not contradictory but complementary. The falling cost of intelligence—likely reflecting cheaper inference and more efficient models—does not automatically translate into consumer revenue. TechCrunch notes that AI remains expensive to operate, and even hundreds of millions of paying customers wouldn't guarantee break-even. The Tom's Hardware report focuses on the supply side, while TechCrunch examines demand. They agree that economics are central to AI's future, but they differ on the trajectory: one sees costs falling sharply, the other sees consumer willingness to pay stuck.
The practical implication is that AI labs continue to pivot toward enterprise, where contracts are more lucrative. OpenAI's enterprise bookings reportedly doubled since July, and even consumer-facing launches like Dots carry an enterprise angle. For startups like Instinct, which plans to take a cut of purchases made through its agent, the path to scale may require either tapping enterprise revenue or betting that falling intelligence costs will eventually make consumer pricing viable. For now, the two reports suggest a market where the technology gets cheaper, but the customer base for consumer AI remains stubbornly narrow.