Not-knowing discussion #13: Tools for thought and action (summary)

30/1/2024 ☼ not-knowingiiiisummary

This is a summary of the thirteenth session in the InterIntellect series on not-knowing, which happened on 18 January 2024, 2000-2200 CET.

Upcoming: “A clear view (?),” 15 Feb 2024, 2000-2200 CET. The last episode of my Interintellect series on not-knowing is a synthesis and recap of this 14-month journey towards a clearer view of not-knowing and how to relate better to it. We’ll focus on practical, concrete, usable learnings. The topics we’ll cover include:

  1. The importance of clearly differentiating between risk and uncertainty,
  2. How being clear about uncertainty makes it easier to do new things and be happy in the midst of change,
  3. The four different types of non-risk uncertainty and what they mean for how we make choices,
  4. What a mindset for not-knowing looks like,
  5. A toolkit for relating well to not-knowing.

First-timers welcomed with enthusiasm! More details and tickets here. As usual, get in touch if you want to come but the $15 ticket price isn’t doable — I can sort you out. And here are some backgrounders on not-knowing from previous episodes.

Tools for thought and action

Reading: Building a toolkit for not-knowing.

tl;dr: The core insight about an appropriate toolkit for not-knowing is that we need tools for different types of not-knowing (about actions, outcomes, causation, and values) but also important are tools for diagnosing situations, updating actions, and building capacity for dealing with not-knowing. Of these, capacity-building tools seem to be the most vital.

Participants: Mo A., Chris B., Trey L., Indy N.

Discussion highlights

  1. Organizations create powerful barriers to relating well to not-knowing. There’s a strong visceral affective reaction against the implications of responding well to not-knowing. This is especially true if the not-knowing implies existential change — this creates avoidance, paralysis, and other forms of inaction. Business history has many examples of this, often interpreted as failure to stay ahead of developing technology (e.g., Kodak and digital photography, Philip Morris and vaping), but maybe better interpreted as internal antibodies” killing attempts to relate to the not-knowing the organization faces.
  2. The barriers to relating to not-knowing seem surmountable or avoidable. There are three possible ways over/around this barrier:
    1. Fear and desperation: There may be a level of fear that motivates an organization to abandon the status quo and engage with not-knowing, but does not tip the organization into panic/paralysis/collapse. This calibrated, intentional level of desperation is what I call desperation by design.
    2. Resilient identity: A strong sense of coherent identity not anchored in subordinate, concrete detail makes it possible to confront and relate with uncertainty. (But successfully superordinating identity and work in this way is hard because it involves intentionally moving up toward more abstract anchors of identity).
    3. Awareness: Cognitive awareness (that formal risk isn’t everything) and standard operating procedures (that call attention to non-risk forms of not-knowing) may override the affective barrier to relating to not-knowing.
  3. Organizations need champions to attend to uncertainty that isn’t risk. But it is an open question as to what the job of Chief Uncertainty Officer (CUO) should look like. Some considerations in designing the role:
    1. Scope and complementarity: A CUO should be responsible for both mitigating fallout from uncertainty and promoting structures that use uncertainty to enable innovation and adaptation. The CUOs role is thus different from that of the Chief Risk Officer, who is responsible for controlling and mitigating fallout from risk but not for innovation and adaptation. The CUOs role complements the CROs and the Chief Innovation Officer’s roles.
    2. Existing analogous roles in corporations: Shell and Salesforce (among others, including governments and militaries) have Chief Futures Officers and futures/scenario planning functions. Futures originally functioned as a cross-organization center for questions and problems that quantitative risk management was incapable of answering. But the ambit of futures groups seems to be shrinking, possibly because a) of the growing strength of the more legible quantitative risk management community, and b) the tension between futures recommendations and investor/shareholder desires (e.g., consider Shell and investor pushback on its sustainability strategy).
    3. Non-corporate analogous roles: One of the roles court jesters and shamans played was channeling and managing uncertainty and incomprehensibility for their communities. They were able to do this because the shaman/jester was given power/tolerance by their community/lord and they were partly outsiders.
    4. Uncertainty champions need both enough power and outsider status: New product development (NPD) teams and managers are ostensibly empowered to imagine long-term newness and fail in order to succeed. But they are rarely genuinely able to push back against rest-of-organization inclinations toward legibility, short horizons, and success. Similarly, people or groups performing red team functions (i.e., pretending to be a hostile group attacking the organization’s strategy, tactics, market, products, or infrastructure) are always marginalized eventually and lose influence.
    5. Should the CUO function be performed by board members in conjunction with external consultants? Consultants are free from some of the organizational power dynamics that employees will always be exposed to — but may lack formal influence. The CUO function might be deliverable by combining consultants’ independence with the formal influence of board members who are obliged to treat the organization as an organization in the long term (instead of focusing on current management teams and short-term performance).
  4. Organizational capacity-building is vital. This turns out to be a lot of distinct capacities: a) Detecting, b) correctly recognizing, c) acting in, and d) responding appropriately to, different kinds of not-knowing. Because of the barriers to even engaging with not-knowing highlighted above, initially organizational capacity will consist of first being able to sit with not-knowing and acknowledge its existence (instead of denying or avoiding it), then being able to take small, learning-oriented actions in the context of not-knowing (instead of being paralyzed and unable to act at all or taking inappropriate actions).
  5. Capacity-building tools are therefore vital. Some considerations:
    1. Capacity-building tools already used in corporate contexts. Some examples include idk as a form of exposure therapy for uncertainty, corporate mindfulness programs (like Search Inside Yourself), wargaming/strategic rehearsal, red teaming, and scenario/futures training.
    2. Corporate capacity in relation to not-knowing looks a lot like general psychological capacity to be open-minded and exploratory, though a) there are individual differences in uncertainty tolerance, and b) personal, professional, and organizational uncertainty are all different. Nonetheless, it’s probably fruitful to look for capacity-building tools in other, less conventional contexts too:
      1. Games, including variants of tarot (for exploring tropes) and role-playing games.
      2. Intentionally fictive exploration, including design fictions.
      3. Chemically-/process-altered states. Nootropics and cognition-enhancing drugs are already semi-normalized in some professional settings (e.g., Ritalin/Adderall in academia, or cocaine in finance), and psychedelics and depressants for some types of creative endeavours (e.g., alcohol for writing). Systematic, intensive and extensive programs — analogous to MKUltra for uncertainty, or Pervitin use by the Wehrmacht — are probably far from being normalized at the moment.
      4. Self-reliance training, such as going on wilderness guiding courses or using adventure playgrounds (Abenteuerspielplätze).
  6. Better questioning is a meta-skill (or meta-capacity) that’s important for relating well to not-knowing. Why? Because questioning how the situation should be interpreted (“Is it risky or truly uncertain?” and If there is true uncertainty, what types of not-knowing are we facing?”) is the first step to relating well to not-knowing. In fact, questioning is a generally valid approach to dealing with not-knowing. (Here, questioning” covers stuff like problem-definition, hypothesis design, and experimentation.)
  7. But organizations rarely operationalize questioning well. Apparently, organizations are not as good at experimenting as they want to be. 1. This is because organizations lack questioning sophistication. Managers often talk about the importance of well-defined problems or good experiments, but generally there is poor conceptual understanding of or practical insight into how to develop good problems, clear hypotheses, and effective experiments. 2. Questioning is therefore underutilized even when it is overtly encouraged. Instead of experimentation and investigation, organizations tend to instinctively revert to the risk mindset and collapse down to clearly legible decisions.
  8. Fortunately, better questioning culture and practice is also trainable. This is doable simply by deciding to do it! 1. Define and teach explicit sensitivity to question quality. Internally, define what makes a question/hypothesis/experiment good. (This is necessarily a subjective question.) Then provide employees with training to recognize and be able to articulate this quality. Workshops in which groups of employees work together on designing business-relevant questions (or hypotheses or experiments) are valuable for this; framings like foundationality (which was recently deployed at SCAI), eigenquestions, and hypothesis-driven development (among others) can help. 2. Teach about different types of experiments. Experiments are designed differently depending on what they are intended to achieve. A non-exhaustive list of types of experiments: 1. For calibrating an understood mechanism. 2. For systematic learning. 3. Individual experiments vs portfolios of experiments. (More here on portfolios of experiments for responding to not-knowing.) 4. (A portfolio of experiments) for actions research. 5. (A portfolio of experiments) for outcomes research.
  1. Scrum poker as a planning tool for agile projects.
  2. Cross-team collaboration on trade-offs when building AI products (an early article draft about diagnostic tools, for comment).
  3. Using tabletop games for therapy.
  4. What is the right amount of strategic ambiguity? 
  5. Iboga, the drug that makes you feel dead.
  6. Economic downturns appear correlated with emergence of new consulting frameworks.
  7. Using checklists as a way to manage well-understood risks.
  8. A field guide for rapid experimentation.