Book Review: Farsighted by Steven Johnson

I work as part of a corporate strategy team. Our focus and our group name is Strategic Foresight. Whereas our product development teams look five years into the future when putting together our product roadmaps, our team tries to look ahead even further. We imagine how our customers will work a decade from now and work backward to see how our software can get our customers to that future. Alternatively, we look at what our software does now, examine current trends (e.g., machine learning) and envision how our software could evolve going forward over a decade. It's a challenging job as the future is hard to predict.

One of the things we do is try to be scientific about our jobs by systematically applying strategic foresight techniques like backchaining and forward chaining. It was in the interest of such science that I read Farsighted: How We Make the Decisions That Matter the Most by Steven Johnson. There is a relationship between decision-making and imagining how our customers will work in the future because good decision-making is based on recognizing all of the possible outcomes of a decision. Predicting how our customers will work in the future involves recognizing all of the possible outcomes associated with changes in society, technology, the economy, the environment, and politics.

Here are some notes I jotted down while reading the book.

• Moral Algebra

  In the introduction, Johnson explains why decision making (when the outcome is important) is challenging.

  It's because complex decisions:

  • Involve multiple variables.
  • Require full-spectrum analysis.
  • Force us to predict the future.
  • Involve varying levels of uncertainty.
  • Often involve conflicting objectives.
  • Harbor undiscovered options.
  • Are prone to being made by the intuitive, fast-acting, emotionally charged part of the brain instead of consciously thinking through a situation.
  • Are vulnerable to failures of collective intelligence.

  The introduction gets its title from a letter from Benjamin Franklin where he described a decision-making process of listing pros and cons on a sheet of paper and canceling out items of equal value as moral algebra.

• Mapping

  Some highlights from this chapter include:

  • Loss aversion refers to the phenomenon where humans are more likely to resist losses than to seek gains.
  • Anchoring is what humans do when given a set of multiple, independent variables related to a situation, they select one "anchor" variable and base their decision on that.
  • Influence diagrams can cover the full spectrum of factors that are related to a decision and help decision makers see the chain of events (called impact pathways) that stem from a decision.
  • In pursuit of covering the full spectrum related to a decision, consulting a wide range of stakeholders allows decision makers to get a variety of perspectives on a problem at hand.
  • A design charrette is an open, deliberative process where small groups of diverse stakeholders critique an existing plan or suggest new ideas to resolve the problem on hand.
  • Although diversity of participants in decision making improves the quality of the decisions made, it also reduces the confidence the participants have in the correctness of their decisions. Decision makers are more aware of what they don't know.
  • Asking providers to apply confidence levels to information they're supplying allows recipients to know how seriously to take the information but also forces providers to consider what they might be missing.
  • When faced with a "whether or not" decision, the best decisions normally result from expanding the set of available options to more than two to chose from.

  Regardless of whether it is physical or metaphorical, no one makes a difficult or important decision without some kind of map.

• Predicting

  It's impossible to predict the future, but there are techniques that can improve one's ability to be correct.

  • Those who are open to a wide range of potential sources and willing to admit uncertainty, not devoted to some overarching theory, are better at predicting future events than single-minded experts.
  • The fallacy of extrapolation is that an identified trend will always continue in the same manner.
  • Randomized controlled trials using large sample sizes and randomly selected control groups revolutionized the predictability of how treatments would affect illnesses. Such trials can be applied to many situations other than medicine.
  • Scenario planning hones in on uncertainties related to a decision and forces those involved to imagine multiple versions of how a decision may unfold. Scenario planners often build one model where things get better, one where they get worse, and even one where they get weird.
  • Simulations are a powerful tool because they allow one to make useful predictions about the future without having to completely understand how a system works.
  • A premortem is an exercise where one imagines that a decision has been made but failed. The exercise is to determine why. This is done before the decision has been made as a way to conduct a full-spectrum analysis to avoid the fallacy of extrapolation or overconfidence from confirmation bias.
  • Given a decision and imagined outcomes, red teams explore alternative narratives. By taking on a separate red team identity, participants gain a new perspective and are more effective than traditional team members playing devil's advocate.

  Accurately predicting the result of a decision stands less of a chance if all of the possible outcomes have not been considered.

• Deciding

  Decisions often have objective and subjective aspects to them.

  • Jeremy Bentham: "Moral decisions — both public and private — should be based on actions that produced the greatest happiness for the greatest number."
  • Many decisions are made using a cost-benefit analysis, but there are often intangibles that should be factored into the decision-making process as well.
  • Value modeling is essentially a weighted analysis. Individual elements related to a choice are scored. Each category of elements related to the decision is weighted. The weights and scores are combined to produce one number per choice. The choice with the best value is the one to male.
  • A Bad Events Table is a precomputed value model that aids in decision-making for technologies like autonomous vehicles.
  • Independent of all of the mathematical ways to approach decision-making, there is no substitute for mulling over a decision to gain additional insight.

  Like most things in life, good decision making is a combination of art and science.

• The Global Choice

  According to Star Trek, the needs of the many outweigh the needs of the few or the one. But do they always?

  • Today we have the capacity to consider and make decisions whose timescales would have been unimaginable to our grandparents.
  • Out of fear of being conquered by a superior civilization, in 2015, a group of science and technology luminaries that included Elon Musk signed a statement that recommended that we not attempt to communicate with life-forms living on other planets.
  • Many are familiar with SETI — the search for extraterrestrial intelligence. METI is the messaging extraterrestrial intelligence. Attempts at METI would not yield any results for 100,000 years. Given its profound implications and long timeframe, who decides if we embark on METI?
  • Moore's law and machine learning have created the possibility for a superintelligence that outperforms humans. Will future decisions be delegated to this superintelligence?
  • Formed in 1961, the Drake equation defines the probability of intelligent life existing on other planets, and a big influence on the result is the amount of time a civilization exists after it advances enough to send/receive transmissions to/from outer space before it ceases to exist by running out of resources or extinguishing itself.

  Decisions are often important because of the severity of a bad outcome or the probability of an undesirable result.

• The Personal Choice

  From its title, I should have predicted that this chapter would be a little self-indulgent. I guess Johnson just couldn't resist summarizing the plot of Middlemarch because it demonstrated a full-spectrum analysis and the various levels decision-making should be considered. That plot summary aside, here are some tidbits from this chapter.

  • Charles Darwin's personal decision to publicize On the Origin of Species was a very courageous decision because although he thought he was on to one of the most important scientific ideas of the century, going public meant that he would be ostracized by the church. Darwin's wife, whom he loved more than science, was devout in her faith, and surely this would cause a rift.
  • Novels provide us with information that is much like what we get from simulations — the ability to visualize experiences in great detail.
  • "It is not an accident that so many of these tools and strategies that help us wrestle with complex decisions revolve around storytelling." ^[page 203]
  • Psychologists and cognitive scientists refer to the ability to imagine the subjective life of another person as having a theory of mind.
  • A study conducted by the New School in Manhattan and published in Science magazine found that readers of literary fiction had improved theory of mind skills.

  Those with better theory of mind skills are more likely to accurately predict actual outcomes because they can put themselves in the shoes of multiple stakeholders impacted by a decision.

• We Might As Well Get Good At It

  The title of this chapter is an allusion to Stewart Brand's opening sentence to the 1968 Whole Earth catalog "We are as gods and might as well get good at it." This chapter wraps up the book:

  • Wisdom is a combination of creativity, empathy, resilience, and decision-making.
  • A course in decision-making would include history, moral philosophy, behavioral economics, probability, neurology, computer science, and literature.
  • How to make a decision:
    • Build a full-spectrum map.
    • Design a scenario plan.
    • Conduct a premortem.
    • Build a values model.
    • Construct a Bad Events Table.
    • Share hidden profiles among diverse groups.
    • Recognize the value of measuring uncertainty.
    • Seek out undiscovered options when faced with either/or choices.
    • Learn the importance of other-mindedness that can be cultivated by reading great literature.
  • When running controlled experiments is not possible, storytelling is a substitute.

  Indeed, storytelling helps frame future direction.

Although I started reading this book because of the relationship between identifying possible outcomes and strategic foresight, there is also a relationship between strategic foresight and storytelling. Our team has found that storytelling is an excellent way of reaching a shared understanding of the future between Autodesk employees and Autodesk customers. As an example, years ago, it would have been challenging to explain how generative design worked; however, a story like this would have probably done the trick.

Speaking of storytelling, this book was easy to read. Rather than provide a tutorial on how to make proper decisions, Johnson decided to tell stories and allow the methodology to unfold as part of each story. The book tells tales such as Charles Darwin's decision to get married, President Obama's decision to storm the building to kill Osama bin Laden, and George Washington's failure to defend New York in the summer of 1776, to name just a few.

All in all, I am glad I decided to read it.

Foresight is alive in the lab.