Lego has launched 330 new products in 2026 so far and is one of the few consumer brands genuinely thriving across a cost-of-living crisis. The formula is embarrassingly simple: a physical object with no moving parts that requires human hands, time, and attention to become anything at all. Meanwhile, a Fast Company contributor declares that the AI adoption phase is over. Two years of 'who's using it and how fast' conversations in boardrooms have run their course. Now comes the part where you have to actually do something with it.

The Attention Economy's Paradox: Complexity Sells

What Lego and AI share is that both require sustained attention to produce value. Lego builds are not passive. Neither is serious AI deployment. The difference is that Lego gives you a box, a manual, and a satisfying click when pieces fit. Enterprise AI gives you an API, a migration headache, and a product roadmap that changes quarterly. The EU AI Act, now the world's first comprehensive AI legislation at 113 articles and roughly 50,000 words, is the regulatory equivalent of Lego's instruction manual: necessary, comprehensive, and slightly daunting before you begin. Accel-backed Keenable just exited stealth with a $26 million seed round to index the web for AI agents, betting that the post-adoption phase needs better search infrastructure, not more models. .

What the Post-Adoption Phase Actually Demands

Scott Belsky's framework of Precision Generative Workflows is the most useful cultural frame here: the organizations that will extract value from AI are not the ones that adopted fastest but the ones that built genuine craft around the tools. That is, ironically, closer to the Lego approach than the enterprise software approach. The instruction manual matters. The patience to follow it matters more. And the companies selling shovels, search indexes, governance protocols, and regulated frameworks, are the ones betting that the adults are finally ready to build something.