Even if AI allows companies to build products 10x faster, customer spending won't scale accordingly. The resulting flood of products competing for a fixed budget will inevitably increase competition, drive down prices, and depress total market revenue.
The most vocal proponents of AI's transformative productivity are the companies selling AI tools. Their narrative focuses on metrics like accelerated code delivery, which benefits their bottom line, rather than on more meaningful business outcomes like actual revenue growth for their customers.
The current market for AI tools and tokens is a temporary boom. Drawing parallels to past tech waves like browsers and mobile, prices will fall dramatically as models commoditize and free alternatives proliferate. This will deflate the revenue streams of today's dominant AI vendors.
Just as companies obsessed over story points during agile transformations, they now focus on superficial AI adoption rates. True value comes from rethinking organizational workflows and bottlenecks, not from mandating tool usage or tracking superficial engagement metrics.
Dramatically increasing feature output via AI creates a new bottleneck: user attention. A user receiving 100 new features per month lacks the time or inclination to discover, learn, and adopt them, meaning most of the accelerated development effort is ultimately wasted.
The premise of agentic software that automatically builds user requests ignores a critical reality: most user suggestions are bad ideas. This approach bypasses essential product curation and strategy, leading to bloated, incoherent, and potentially harmful products that are difficult to support.
As AI automates the tactical middle of product management—specs, schedules, and engineering liaison—the role will bifurcate. PMs must add value at the fuzzy front-end (discovery, strategy, economics) and the go-to-market back-end (sales enablement, marketing), creating a barbell-shaped time distribution.
When executives demand 10x product output due to AI, product leaders must reframe the discussion around revenue, not code. Challenge sales and marketing teams to formally commit to higher quotas and lead targets for these new products, shifting focus from output metrics to business outcomes.
Product managers are increasingly acting as 'product builders' who code with AI. This isn't a strategic evolution of the role, but a rational response to misguided executive incentives that reward visible, technical output and threaten to eliminate PMs who don't demonstrate these skills.