Not-knowing discussion #12: Broad approaches (summary)

14/1/2024 ☼ not-knowingiiiisummary

This is a summary of the twelfth session in the InterIntellect series on not-knowing, which happened on 21 December 2023, 1700-1900 CET.

Upcoming: “Tools for thought and action,” 18 Jan 2024, 2000-2200 CET. Episode 13 of my Interintellect series on not-knowing is about the tools we need to relate well to not-knowing. We’ll take the broad approaches from Episode 12 and unpack them into some concrete tools like superordination during goal-setting, clearly identifying non-goals, using forced choice to reveal tradeoffs, desperation by design, and implication analysis (among others). 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.

Broad approaches

Reading: Broad approaches suitable for not-knowing.

tl;dr: A mindset for not-knowing helps us avoid sleepwalking into using the same old risk-based approaches for dealing with partial knowledge. Instead, this mindset brings clarity about the variety of non-risk types of not-knowing and how each type arises. This, in turn highlights how each type demands a different broad approach to taking action.

Participants: Nadim C., Paul M., Razan B., Stefan L., Trey L.

Discussion highlights

  1. There are four types of not-knowing (explainers for each type are linked below). Each demands different broad approaches to action.
    1. Not-knowing about actions: Many small experiments intentionally designed as a portfolio to explore the space of possible actions.
    2. Not-knowing about outcomes: Many small experiments intentionally designed as a portfolio to explore the space of possible outcomes, and structured forms of exploratory imagination.
    3. Causal not-knowing: Conventionally understood experimentation and stacking the deck.”
    4. Not-knowing about values: Reasoning/negotiation/imagination about value and goal super-ordination.
  2. These broad approaches are summarised in this figure: Different broad approaches for different types of not-knowing.
  3. Current approaches to action under uncertainty are problematic or incomplete. However, they may be useful as components of a broader toolkit for dealing with different, clearly articulated types of not-knowing. Some of these include:
    1. The Cynefin framework seems to focus on causal not-knowing, and may conflate uncertainty with risk.
    2. Trusting the process” seems to focus on causal not-knowing, and is limited to situations where available actions are relatively well-understood.
    3. Structured imaginary exploration (e.g serious play) seems to focus on not-knowing about outcomes and possibly not-knowing about values, by exploring outcomes and their desirability in low-stakes, play-focused settings.
    4. Backcasting seems to focus on exploring possible actions and causation given known/assumed outcomes.
  4. There may be a progression or telos of development/maturity in recognising and relating to not-knowing. This is a kind of adaptive process that begins with recognising the phenomenon (i.e., seeing the distinction between risk and uncertainty, then seeing distinctions between the different types of non-risk not-knowing contained in uncertainty), then with interrogating contingent approaches for relating to different types of not-knowing).
  5. This series of investigations about not-knowing may capture this progression (!) from more detailed recognition of not-knowing to developing a nuanced toolkit for relating to not-knowing. This article provides a bit more background on project’s genesis and arc.
  6. Many of the broad approaches listed above are alive in the wild.
    1. UNDPs move to project portfolios.
    2. Bureaucrats and politicians using deck-stacking.
    3. Use of speculative fiction in futures/scenario work.
    4. Stakeholder research on relative valuess.
  7. Deploying these broad approaches effectively will require good calibration between abstractness and concreteness, but learning how to calibrate well is hard. Heuristics from different domains may shed light on ways of learning good calibration, particularly opponent-process theory (psychology) and selection on variation (evolutionary biology). Both negative and positive feedback mechanisms seem crucial to achieving good calibration.

Fragmentary ideas/questions that came up which seem valuable

  1. In causal not-knowing, the most important thing to do is interrogate implicit assumptions, but this is particularly hard. What methods are there for recognising implicit assumptions. (Imaginary ethnography is good for this.)
  2. Where there is not-knowing about outcomes and values, setting goals at the right” level of abstraction prevents artificially limiting ways of being successful.
  1. Lego’s serious play..
  2. Using opponent-processing to optimize (vs maximize) in dynamic systems.
  3. Researchers studying dishonesty may have been dishonest themselves.
  4. Stacking the deck as a strategy for managing bureaucracies.
  5. UNDPs shift from individual projects to portfolios of projects.
  6. Portfolios of actions for system transformation.
  7. How fitness in dynamic systems evolves under pressure from both selection and variation.