What Adobe v ArtCraft teaches us

The cost of creating custom software drops drastically, making it feasible for individuals to build applications that were once the exclusive domain of large companies.

Large AI model providers are facing a fundamental challenge to their business models, which were predicated on replacing expensive software engineers with AI. The core issue is that these AI services, often priced as low as $20 per month, are enabling individuals and small teams to rapidly create software that previously required significant engineering resources and capital. This effectively dismantles the "toll booth" model, where companies like Adobe historically profited by making difficult-to-replicate software, thereby charging recurring fees. By democratizing software creation through cheap AI, these labs are eroding the customer base that would have paid premium prices for their AI engineering services. While some companies are incurring substantial infrastructure costs and selling services at a loss, the proliferation of open-source models further exacerbates this problem, making it impossible to maintain premium pricing. This situation mirrors historical economic events like the influx of Spanish silver or the advent of lab-grown diamonds, where abundant supply devalued a previously scarce and valuable commodity.

The AI industry's strategy of offering low-cost AI development tools has inadvertently devalued the very software products and services it aimed to compete with. By making complex software creation accessible and affordable, AI labs are dismantling the traditional business models of software companies that relied on high development barriers and recurring subscription fees. This democratization of creation means that the "toll booth" revenue streams that funded these software companies, and were intended to be captured by AI providers, are rapidly evaporating.

The current pricing strategy is unsustainable, as AI labs are essentially subsidizing the destruction of their potential customer base. The massive infrastructure investments in AI data centers are not yielding profitable returns because the market for AI-generated software is collapsing under the weight of free or extremely low-cost alternatives. This forces AI labs into a difficult choice: maintain low prices and bleed cash while their subsidized products undermine their customers, or raise prices significantly, which would likely alienate users and still struggle against free open-source models.