CPO Summit 2026: 7 Ideas from San Francisco
This week I attended the Chief Product Officer Summit in San Francisco.
I shared my own reflections as a CIO in When Execution Becomes Cheap, Judgment Becomes Valuable.
This post is different: it collects the ideas I heard from speakers and peers throughout the day. Some confirmed what I was already thinking. Others challenged it.
These are the 7 that stayed with me.
1. Execution is getting cheaper. The why is not.
AI is dramatically reducing the effort required to research, analyze, prototype and build. The question is moving from “Can we build it?” to “Should we build it?”
Velocity needs a vector: moving faster only helps if you are moving in the right direction.
One of my favorite concepts of the day: the CPO is becoming the “Chief Why Officer”. What are we building? Why does it matter? What outcome are we trying to achieve?
2. AI doesn’t remove bottlenecks. It moves them.
Product and Engineering may become dramatically faster while Legal, Finance, Go to Market or decision making remain unchanged.
One discussion suggested looking beyond idea to launch and thinking about idea to revenue.
Optimizing one part of the organization is not enough.
3. There is no single AI-first team
A moonshot, a 0-to-1 product and a mature product at scale need different structures.
Some problems benefit from small, senior and highly autonomous teams. Others require specialization, governance and quality controls.
Design the team around the problem, not the trend.
4. Boundaries are blurring, so guardrails matter more
Product managers can prototype. Engineers can access customer insights. Business teams can build workflows and agents.
If everyone can build and agents can act, organizations need clarity: Who owns the outcome? What can an agent do autonomously? When is human approval required?
The goal is not choosing between control and autonomy. It is creating autonomy with guardrails.
5. AI needs company context
Models know a lot. But they don’t automatically understand your customers, processes, priorities, business rules or what your data means.
The advantage may not come from having the best model, but from giving AI the best company context.
6. Don’t automate the training ground
The most thought-provoking idea of the day.
AI is automating many execution-heavy junior tasks. But those tasks are also where people learn by doing, making mistakes and building experience.
One speaker called the potential consequence an “expertise drought.”
We should be careful not to solve a short-term efficiency problem while creating a long-term talent problem.
7. Curiosity and judgment become more valuable
Our roundtable ended with two skills that may become increasingly important: curiosity and judgment.
Curiosity pushes us to explore. Judgment helps us decide what matters.
And judgment comes from experience, mistakes and learning.
Where this leaves me
Many of these ideas point in the same direction: as AI makes execution cheaper, the real questions move to direction, context and people.
For technology and product leaders, the challenge is becoming less about what AI can do and more about deciding where it should take us.
What that means from a CIO perspective, and why the view from Barcelona looks different from San Francisco, is what I explored in When Execution Becomes Cheap, Judgment Becomes Valuable.
