These reading notes gather seven short pieces on innovation and uncertainty. They draw on McGrath, Rumelt, Varol, Viki, O'Reilly, Kay and King, Miller and colleagues, Ahrens, McKelvey and Paul Millerd. They are adapted from David Drysdale's Moby Diction notes.
1Living with uncertainty
1.1Brains try to minimize surprise
Neuroscientists who support "predictive processing" argue that brains are built to minimize average surprise, or prediction error. Our curiosity fits this. It leads us to probe, to discover, to find areas of uncertainty and work out how to resolve them. Miller, Nave, Deane and Clark make the case in an Aeon essay. But uncertainty comes in different kinds.
1.2Three kinds of uncertainty
Naming the kind helps us deal with it.
- Expected uncertainty is a task-relevant gap in knowledge. We know the gap exists and can act to fill it.
- Unexpected uncertainty appears when a change in the environment forces us to question our model. The rules seem to change without warning. We learn faster, which can help us break out of old patterns of thinking.
- Volatility means the rate of change is itself changing. We are uncertain about uncertainty. We cannot be confident that we can learn or master the context, and we cannot tell what the rules are.
John Kay and Mervyn King draw a similar line in Radical Uncertainty. "Resolvable uncertainty" covers situations where you can look up the answer or use a known probability. "Radical uncertainty" has no clear resolution state, and probabilistic thinking alone cannot settle it. It can end in results we did not expect and results we could not have imagined. Resolvable uncertainty looks like expected uncertainty in Miller's terms. Radical uncertainty could cover both unexpected uncertainty and volatility.
1.3Narratives help us act without probabilities
Narratives are not just deception. They are an important tool in decisions where probability judgments are not possible because of uncertainty. We use metaphor, analogy and other language tools to understand a situation and to make the abstract concrete. They answer the question "what is going on here?" We need narratives to make sense of facts, even when numbers are available.
Two cautions from Kay and King:
- No single narrative applies everywhere. Our reference narratives need continual revision as new information arrives. People want a grand narrative for any complex situation. But domains of radical uncertainty are non-stationary and complex. They are always changing, and they may change in response to the narratives we tell about them.
- "Narrative" covers more than stories. It includes models and other tools for thought that imagine fictional agents acting under hypothetical conditions.
To work, a narrative must be credible and coherent. It does not have to be true.
2Betting small
2.1De-risk innovation with smaller bets
Organizations should keep a balance between two kinds of decisions. Type 1 decisions are big, significant and risky. Type 2 decisions are low risk and reversible, and they teach something. At Amazon, small teams can make and act on type 2 decisions without heavy bureaucratic control, which gives the company a steady stream of new insight.
Many organizations make type 1 decisions on large assumptions that may not hold. When information comes back and challenges those assumptions, they fall into the sunk cost fallacy. They will not give up the capital already spent, so they double down until it is too late to save the project.
The better approach is a steady series of small type 2 bets and learning from the results. Several cheap ideas in parallel let an organization invalidate more ideas, faster. Each round removes non-viable options and makes the solution stronger. This also lowers the stakes. When we treat ideas as experiments, we get used to seeing them as working hypotheses, not solutions.
Richard Rumelt adds "proximate objectives," goals feasible enough for the organization to achieve. He suggests they help organizations move through uncertain areas.
2.2Innovation cascades
Innovation begets innovation. When we innovate in one area, we are often pushed to innovate in others. Jim McKelvey describes a nested series of consequences. We innovate in one space, "so we have to" add another innovation to support it. We want to achieve x, and to do that we have to achieve y. Plan for the chain, not for a single change.
3People and judgment
3.1The Tom Brady principle
Paul Millerd's essay questions the habit of promoting top performers into management. Imagine the New England Patriots offering quarterback Tom Brady, after a Super Bowl win and an MVP award, a "promotion" to general manager.
Instead of promoting top performers, organizations should pay people for performance and contribution regardless of title. Success and climbing the ladder should be decoupled. There should be paths for people who excel at leadership and coaching and for expert contributors. Authority is not performance, and a promotion is not the best way to unlock creativity and innovation.
3.2Honing intuition
Sönke Ahrens calls intuition the "sedimented experience on which we build our conscious, explicit knowledge." To intuit good ideas, we first have to consume good ideas. Daniel Kahneman argues that we can hone intuition by acquiring and mastering knowledge. Knowledge earns compounding returns. Each idea connects to several others, and our constellations of thought grow fast.
- Read broadly. Narrow focus gives tunnel vision and hides signals from adjacent areas. A strategist must see old things in new ways, and diverse sources give new vantage points. The near-futurist Rohit Bhargava reads magazines written for audiences far from his own experience. He says that helps him spot trends he would otherwise miss.
- Process what you take in. Hubert Dreyfus says memorizing facts is only the first step. As skill grows we move past memorized rules and recognize patterns. Then we filter good information from noise, then see how it connects as a whole. At real expertise we can intuit a decision without conscious analysis.
4What to take from it
- Name the kind of uncertainty before choosing a response.
- Make most bets small and reversible, and keep big irreversible bets rare.
- Expect one innovation to force others, and budget for the chain.
- Keep narratives and models under constant revision.
- Pay and recognize experts without forcing them into management.
- Build intuition on purpose: read widely and work the material over.
5Sources
- Miller, Mark, Kathryn Nave, George Deane and Andy Clark. "Use Uncertainty to Leverage the Power of Your Predictive Brain." Aeon.
- Kay, John, and Mervyn King. Radical Uncertainty: Decision-Making Beyond the Numbers. W. W. Norton, 2020.
- McGrath, Rita Gunther. Seeing around Corners. Houghton Mifflin Harcourt, 2019.
- Rumelt, Richard. Good Strategy Bad Strategy. Currency, 2011.
- Varol, Ozan. Think Like a Rocket Scientist. PublicAffairs, 2020.
- Viki, Tendayi. Pirates in the Navy. Unbound, 2020.
- O'Reilly, Barry. "Optimize to Be Wrong, Not Right." April 6, 2017.
- McKelvey, Jim. The Innovation Stack. Portfolio, 2020.
- Millerd, Paul. "The Tom Brady Principle: Don't Promote Your Best People." Boundless: Beyond The Default Path, November 4, 2018.
- Ahrens, Sönke. How to Take Smart Notes. 2017.
- Kahneman, Daniel. Thinking, Fast and Slow. Anchor Canada, 2013.
- Bhargava, Rohit. Non Obvious Megatrends. Ideapress Publishing, 2020.
- Dreyfus, S. E., and Hubert Dreyfus. "A Five-Stage Model of the Mental Activities Involved in Directed Skill Acquisition." 1980.
- Madsbjerg, Christian. Sensemaking. Hachette Books, 2017.
Adapted from David Drysdale's Moby Diction notes.