trust the process

To innovate successfully we must command uncertainty. Before you can command something, you must first notice it. Action towards an (uncertain) goal produces (uncertain) evidence about that goal which is useful for deciding the next action. This iteration can spell the difference between indecision and relative clarity. Clarity on whether to pursue or discard a goal is enormously valuable. In this simplistic framework, clarity emerges when we align on an (uncertain) goal, take excellent action and collect high-quality evidence. The belief about the goal prior to taking action is called just that. The updated belief after evidence is called a posterior.

The mechanics of this process are obvious in the context of an individual, less so for an organisation. How do I know you calculated the posterior correctly? How do we retain the details when communicating posteriors or joining them together? How do we keep each individual posterior in sync and consistent across individuals and time? How to avoid sounding annoying while saying "posterior" all the time?

"You show me a pay stub for $72,000, I quit my job right now and I work for you." - Donnie Azoff, Wolf of Wall Street, 2013

Great teams do great work and great teams are built on trust. You're a sucker if you trust anyone blindly. You're a sucker squared if you feel entitled to others' trust. Trust is mutually earned. In the ancient times, strangers in the workplace built trust by weathering thousands of shared decisions, responsibilities, successes and mistakes over many months. In the age of AI, we no longer have to trouble ourselves. Simply have everyone write down what they think the posterior is for this week, let Claude summarise the submissions and Jev evaluate the correctness. Just kidding.

There is no secret. Since we do not trust each other (yet) and our goal is to always maintain a shared prior and the prior is yesterday's posterior that is calculated from evidence we must - drumroll please - calculate the posteriors together! That is to say: examine the evidence together, see what your first-brain throws up. This sounds ridiculous, you say. I made you listen to all the prior-posterior gobbledygook only to say - just play it as it lays. Realise, nay, trust - people disagree when they lack access to the same evidence. Conversely, smart people are much more likely to agree or at least compromise - the decision just feels right. How something feels does not make it actually correct: evidence may be missing, timings differ, etc; but that does not matter. The iterative process we agreed to follow will get us there, we just need to get to Friday marginally smarter than we started on Monday. The goal is to build good decision-making muscle at every level. The strength of this muscle can be measured by how few decisions leaders have to make day-to-day. But it is felt like trust.

If I had to make a recommendation: share evidence and make decisions together. In the pre-trust stages, curate the evidence religiously: there are libraries filled with books about what this means for your task. Build your posterior, share it, defend it successfully and prepare to discard it if you cannot. Anticipate questions or expect to not be taken seriously. Accept the feedback - you are not your work. Also, chill on the feedback - ask for evidence. Some may call this being "data-driven", presumably in contrast to being ego, fear, status, dogma, hubris, myopia, paranoia, vindictiveness or optics-driven. This is just one suggestion.

If building trust together feels like a faff, it's because it is. Again, trust here is whether in an arbitrary pair of teammates, each believes the other takes valuable action and both know that their actions are more valuable together. A leader can consider an arbitrary collection of things valuable. This collection has a name - culture. It gives the only consistent way to measure performance. Culture needs to be defined, shared, updated, hired and fired for, not written on the careers page and promptly forgotten.

My entire argument prescribes a process. But most process tends to be less structural and more prescriptive. Top-down step-by-step guides may be appropriate for high-risk situations and/or inexperienced teams, but in the long run, they are a hindrance. Process is easier to modify than culture. It's a seductive way to idiot-proof an organisation by curtailing individual judgement, which is doomed to fail when its assumptions expire. Since people tend to agree when all assumptions are present, they can avoid future issues with more context, not more process. The base drive to idiot-proof may point at cultural issues.

"If you want to build a ship, don't drum up the people to gather wood, divide the work, and give orders. Instead, teach them to yearn for the vast and endless sea."

Prescribing a format, a mechanism or a system risks establishing top-down process bound to collapse under its own assumptions. Under culture-first framing, drift and forgetting become a tactical rather than strategic concern. Strategic concerns rest with the leadership, while tactics are situational and free to change as long as they serve the strategic objectives.