Why there's nothing more practical than a good theory
Ah, that old saw again, “I agree in theory, but is this practical?”
This reticence is precisely what our company is targeting: That the sheer uncertainty of how AI will evolve makes deciding on how to plan for its adoption very difficult. Most clients I speak with are heavily focused on driving initial adoption. Getting AI development harnesses into the hands of their developers and beginning to work with their business colleagues on the use of generative AI to replace laborious or mundane automatable workflows. However, looming over them is the realization that there is a revolution under way that will reshape the very business and its use of technology far beyond what is currently being rolled out today. And the fact is that nobody knows what the future operating models and future industry dynamics will be in 18 months, let alone three years.
A book that transformed my thinking in this regard was written over 20 years ago by Peter Block, “The Answer to How is Yes”. The big idea behind this book was that while leaders obsess over method, approach and planning it is often an excuse for inaction. It is a, not necessarily unwarranted, attempt to exert control in the face of anxiety about change. I am sure we have all seen it and done it. Accepting somebody else’s ‘how’ puts a great deal of trust in them, their assumptions and agenda. When this is a person inside your organisation that may have political or organisational implications for the leader too. So, we hire consultants to provide a ‘neutral’ view. It is obviously not neutral and is heavily swayed by the sponsor’s direction. Whatever the reason, we justify our own inaction.
When planning this little essay, my intention was to talk about how the best way to mitigate those concerns was through small, safe experiments. Since the early days of cloud the best way to test out new technology was this. It puts little at risk operationally or personally. However, I think that we need to address the elephant in the room. What’s in it for me to take any risk at all, at any scale? Why do it? Why not just go with the flow - let others test the big ideas and be seen as the sensible fast follower?
Strategy, for any leader, is as much a representation of personal values as it is commercial value.
Getting clear on ‘purpose’ is key according to Block and should precede consideration of approach (how). Focusing on what your values are and how to use your personal agency to drive them through everything you do is the starting point. If this is not the purpose of your personal agency then what is? Now this may sound rather idealistic, but in the end we all have to live in the organisations we build. So really, the question for the leader, whether team or division or organisation-wide, is ‘what do you want your organisation to look like?’.
As engineers and architects, we often frame AI in terms of what it might take from us: decision-making, agency, jobs, careers. These are practical concerns, but they leave a more important question unanswered: what could we do with our agency if we chose to use it differently? What is it currently stopping us from doing? Answers often lead back to familiar technical ambitions, many of which we have held for years: automate a process, codify a policy or redesign a service. AI makes these things easier, it doesn’t stop us from doing them. It certainly changes our relationship to how we deliver them; we specify and design rather than actually write the code.
So what remains is, “what do we want to (or maybe, what should we) do now that AI makes the obvious easier?” That comes down to purpose, as Block says. It is what AI is not able to do for us. We spend a great deal of time talking of ‘loss’ due to AI but seem unconcerned about filling that gap with new potential.
Dromologue provides a mechanism for defining how the organisation will function; interact with other organisations, produce value, structure itself and manage risk. But our clients need to provide the purpose. We can help them articulate that using a variety of approaches. And when time comes to test out those ideas, the answer to how is yes, pick a few teams and just experiment in an initially small and safe way. Dromologue provides the software to guide and monitor and report on the impact of the changes made. Beyond the anxiety that this may entail, there is nothing more practical than real live data on its impact to decide what to do next. But you need to have an initial theory or hypothesis as to how you want to change your organisation and for what effect.
There’s nothing more practical than a good theory. And a good theory can always be tested.