AI is changing the game
It even has people asking if we still need Scrum Masters, and if Scrum itself is still relevant.
Mechanical implementations of Scrum, which merely orchestrate people and work, are under pressure.
After all, AI can radically accelerate development without the need for cadences and formal events.
AI can help generate product visions. It can summarise conversations, analyse data, critique, advise, and generate ideas. It can create and refine Product Backlog items. It can orchestrate, develop, and test.
However, AI also introduces new challenges that need human oversight and there are still things it cannot do alone.
It can guide, for instance, but it can also be ignored or not be consulted in the first place.
It cannot be the unignorable, present, aware, psychologically sensitive, provocative change agent in the room.
It also cannot remove uncertainty.
In fact, the acceleration driven by AI increases uncertainty.
That means there’s one critical need that doesn’t go away. In fact, the need for it is more acute. Empiricism.
Creating products faster is great, but if we’re not careful, it could be faster in the wrong direction!
Let’s remember why Scrum emerged
Scrum didn’t emerge because Developers were slow at coding or because they lacked the ability to organise themselves.
They might not have had the autonomy to organise themselves, but that’s a different question, and one of the reasons for an active change agent in the room.
Scrum emerged because managing work across silos multiplies waste, delay, and unpredictability, while eroding customer centricity, ownership of outcomes, and quality.
It emerged because decision latency was a serious impediment to innovation.
It emerged because predictive planning fails to predict what will happen and when in complex work; or how needs and circumstances will change.
It emerged because users and customers need useable value now, not in three years’ time.
Scrum emerged because continuous adaptive planning is the only resilient way to navigate complexity.
It’s not just that we can’t predict how circumstances will change, or what competitors will do that disrupts our plans, or how needs will organically change over time.
It’s also that what people want changes when they have a product in their hands.
Value creation itself changes the game.
The move from predictive frameworks to adaptive frameworks was not an aesthetic choice. It was a necessary response to this realisation.
The new challenges AI presents
For all its benefits, AI can have issues with hallucination, confabulation, fabrication, and error.
It is also changing the shape of ethical issues such as privacy.
AI needs skilled, AI-aware oversight.
A deeper look at acceleration and uncertainty
Let’s look at the cone of uncertainty, a way of visualising uncertainty. The more we look ahead in time from where we are now, the more uncertain things get.
Here’s what’s happening with AI.
We are testing our hypotheses faster than ever before.
We are uncovering problems faster than ever before.
We are invalidating assumptions faster than ever before.
We are innovating faster than ever before.
As the speed of innovation increases, the cone of uncertainty widens. The horizon of relative predictability gets shorter.
This means our feedback loops need to get shorter.
This is why our inspection game needs to get more acute.
This is why our adaptation game needs to be more effective.
We need to raise our empiricism game, and AI can help us do it
Now is the time to remember that Scrum was never about pure speed or ceremony. It is about maximising quality and value through empirical process control.
It is also about changing the culture of teams and organisations.
Now isn’t the time to stop practicing empiricism. It’s time to raise the game — and the present, aware, provocative change agent in the room is still needed.
However, if we’re not using AI to do it, we’re missing a trick.
A huge trick.
What does this mean for Scrum Masters?
It means we must lean into AI.
It means we must raise the empiricism game.
It means we must ensure oversight of AI, and keep the Scrum Values, and the principles of Transparency, Inspection, and Adaptation.
It means we must continue to be the active change agent in the room.
Get started with the Professional Scrum Master — AI Essentials training, which I am teaching myself on May 12.