Disruption and Denial

Marty Cagan argues that the current resistance to AI products mirrors the denial that greeted the Internet in the mid-1990s. Back then, many organizations insisted their old Waterfall funding, building, and shipping habits still worked, and dismissed cloud architecture as unsuitable for sensitive data. Cagan sees the same pattern now: people say AI is just another feature and that probabilistic outputs and hallucinations rule it out. He contends that discovering and delivering intelligent products differs substantially, and that team topology, roles, and discovery practices will need to change. He expects startups to move fast while large enterprises with much to lose will resist, until a competitor shows customers a dramatically better solution.

01Key takeaways

• Expect denial when a new enabling technology arrives; many people will defend existing roles and processes.
• Recognize that AI-driven products require different discovery and delivery approaches, not just an added feature.
• Plan for probabilistic behavior by using proven techniques that mitigate hallucination risks rather than avoiding AI entirely.
• Start building intelligent products now, since waiting risks being disrupted by a competitor with a dramatically better solution.
• Watch for shifts in product team topology and roles as AI changes how teams discover and deliver solutions.

02Key sections

The Internet objections of the 1990s

Cagan recalls evangelizing the Internet as a platform at Netscape and hearing that old roles and Waterfall methods still sufficed, and that cloud storage was too risky for sensitive data. These objections motivated his first edition of INSPIRED.

The same denial with AI

Today's most common objection is that AI is only a new feature and product discovery and delivery remain unchanged. A second objection is that probabilistic systems that hallucinate cannot be trusted or fully tested in advance.

Substantial changes ahead

Cagan argues that discovering and delivering intelligent products differs meaningfully, and expects shifts in product team topology, team roles, and how solutions are discovered and delivered.

Mitigating the risks

He acknowledges hallucination risks but says strong product teams can use appropriate techniques to mitigate them, as covered in his piece on Creating Intelligent Products.

Startups versus incumbents

Startups will move quickly to seize the opportunity while large enterprises resist, until a new competitor disrupts them. He illustrates the pattern with the salesforce.com story in a footnote.

03From the post

“A partnership dedicated to teaching best practices to product teams and product leaders”

04Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.