AI Moves Things Faster. But in Which Direction?
What I noticed about Agency, Career Lattices, & Why Speed Without Conviction Is Just Drift
I went for a walk this week with Salvador Restelli, Director at Mobile.de and Productlab Conference board member. We ended up talking about agency. Not the buzzword version. The real one: how people, teams, and companies take strategic bets, build authority, and push organizations in a direction that matters.
Salvador reflected on his own career and the moments where agency was missing. Not because the skills weren’t there. Because the conviction to push wasn’t. Waiting for the organisation to align itself. Time passed, People moved on, Initiatives (the expensive ones) became an un-read Google Drive file.
Dara Khosrowshahi said it plainly on Diary of a CEO.
He took over Uber when the company was losing $3 billion a year. The turnaround didn’t start with a new strategy deck or a product vision. It started with saying the uncomfortable truth before anyone asked for it. Proactively. Not softened, not over-contextualised, not buried in a 40-slide narrative. Just: here is what’s actually happening.
He described pinging people on Saturday. And if needed, on Sunday. Not to micromanage. To get to reality. To make sure the organisation was operating on truth, not on the version of truth that’s comfortable enough to present upward.
That’s agency. And it took Uber from $3 billion in losses to $9 billion in free cash flow.
The thing that struck me isn’t the scale. It’s that the mechanism is so simple and so rare.
Most product leaders I observe do the opposite. They soften findings. They hedge. They add so much context to bad news that it loses its weight. They wait for the quarterly review to surface what they knew in week two.
Saying the hard thing before being asked is not a personality trait. It’s a strategic choice. And the leaders who make it consistently are the ones people stay loyal to.
Diana Stepner wrote about the career lattice this week, and it reinforced the same point from a different angle.
The PMs who reach Head of Product aren’t the ones who waited for promotion. They’re the ones who made moves. Sideways, diagonal, sometimes deliberately down. They mentored someone before it was their job. Led the project nobody owned. Built a track record in the open, before anyone gave them permission.
That’s agency again. Not career optimisation. The willingness to act before the conditions are perfect.
The Group PM player/coach trap is another version of the same story.
Most first-time Group PMs play too much and coach too little. The ratio should be 60-70% coaching, 30-40% hands-on. Almost everyone inverts it.
Why? Because doing the work yourself feels like agency. It feels productive. But it’s actually the opposite. It’s avoiding the harder move: letting go of the thing that made you successful and trusting that developing other people is how you create more impact than you could alone.
Real agency at the leadership level isn’t doing more. It’s deciding what matters, saying it clearly, and building the conditions for other people to move.
Florian Bonnet wrote something on his account this week that tied all of this together for me.
Product judgement is probably the #1 skill for Product Leaders going foreward.
AI moves things faster. But in which direction?
That question is the one I keep coming back to. Because speed without conviction is just drift. You write faster, iterate faster, test faster. But if nobody in the room has the agency to say “we’re building the wrong thing” or “this strategy isn’t working” or “we need to make a different bet”.
Or simpli we are just getting busier on iterate and vibe coding, without real business outcomes?
Then all that speed just means you arrive at the wrong place sooner.
The AI shift is exposing something that was always true but easier to hide: the gap between leaders who have judgment and leaders who have process. When execution was the bottleneck, you could build a career on being reliable, organised, and good at managing the backlog. That’s not enough anymore. The bottleneck has moved to the decision layer. And the decision layer runs on agency: the willingness to take a position, push for it, and be accountable when it’s wrong.
What I’m paying attention to.
I keep noticing the same pattern across conversations for potential conference’s speakers, community events, and the things I read on LinkedIn. The leaders who are building real credibility in the AI space aren’t the ones with the best frameworks or the most polished communication style. They’re the ones who act. Who push to reality instead of waiting for it. Who say what they see before anyone asks. And move organizations that are too slowly adopting AI.
That’s the direction that matters. Not faster. Not more efficient. More honest. More convicted. More willing to be wrong in the open.
AI gives us speed. Agency gives us direction.
I think a lot of product leaders in Europe are about to discover which one they actually have.
With love from Berlin
Daniele
Photo by Serj Sakharovskiy on Unsplash