Reading Blink (2007-09-04/13)

Decisions are difficult. A good decision requires work; a bad decision brings consequences. More information sometimes helps, and--with the Internet--much information is available. However, consultant work gives me this perspective: more research delays a decision and increases its fee. Even if the decision is better as a result, the difference may be marginal and of less value than the lost time. During the delay no action is taken, which may have an opportunity cost. So I "sleep on it", stare out the window, or take a walk. I sometimes use these approaches--and my judgment--instead of additional information.

Malcolm Gladwell filled Blink: The Power of Thinking Without Thinking with anecdotes and studies about the effect of additional information on decisions. Some people assume that additional information always improves thinking; Gladwell cites situations in which that is not the case. An example is classical music auditions, which were biased against women until held behind a screen. The additional visual information actually distracted from evaluating the performance.

In the Afterward, Blink asks "When to Blink--And When to Think." The initial answer comes from Ap Dijksterhuis' studies on "unconscious thought" (which BBC News and New Scientist have summarized): think about simple decisions, sleep on complex ones. However, further reflection leads Gladwell to two qualifications: unconscious thought requires training, and statistical summaries suggest significant factors.

Gladwell gives examples of effect on judgment, both good and bad. Training improves judgment in the battle of Chancellorsville and the Millennium Challenge 2002. Biases have a bad effect on decisions we make in the "blink of an eye"; examples are the shooting of Amadou Diallo and the results of an Implicit Association Test.

In addition to training, statistical summaries also improve decisions by suggesting significant factors. The Wages of Wins suggests it is difficult to estimate summaries from incomplete observations of small differences. A report from Cook County Hospital exemplifies how statistics can isolate the factors relevant to a decision.

Statistical summaries also have their role in my consulting. I just completed a project requiring analysis of millions of records I imported from daily data and queried for mapping and evaluation. I use statistics, training, and unconscious thinking to improve the speed of my decision-making.


Walking to Work (2006-10-26)

Sunrise in Montmartre
Walking to the bus reminds me of my morning commute in Paris and renews my commitment to venture and work abroad again. In Paris Tuesdays through Thursdays we would awaken in Pigalle. Ryan went to school in Saint-Germain-des-Prés and Jean Claude went to work in Montmartre. I followed Jean Claude out onto Rue André Antoine, smelling the wet cobblestones as the street cleaners washed away the previous day. I heard the sounds of other pedestrian's shoes on the wet cobblestones and an occasional car vibrating along the street. The October air cooled me as I leaned into the steep hill and climbed. One morning in late October I took a photograph of the sunrise in Montmartre. Later on those weekdays I walked back down to Le Chao-Ba-Café for the afternoon sunlight.

On weekends--sometimes long weekends including Friday and Monday--Ryan and I would travel. We visited Montpellier, Nantes and Tours, Chamonix, Berlin, and London. We didn't visit Barcelona or Venice like we had hoped.

Memories and daydreams make me think of "Disappearing Act: How to Escape the Office", chapter 12 of The 4-Hour Workweek. A colleague who also develops software for Geographic information systems is currently in Turkey while his wife teaches. Perhaps I can do the same somewhere while Ryan teaches ESL.


Reading The Diamond Age (2007-08-26/09-03)

Neal Stephenson's The Diamond Age: Or, a Young Lady's Illustrated Primer has a future earth setting in which nanotechnology has significantly reduced scarcity but artificial intelligence has not been achieved. I enjoyed Cryptonomicon by Neal Stephenson (especially its encryption themes, which are present as well in The Diamond Age) and decided to read another of his books.

The "Diamond Age" is due to molecular control of matter making diamond--with multiple productive properties--a common material. (Wikipedia attributes this idea to "It's a Small, Small, Small, Small World" by Ralph C. Merkle.) It is interesting that the social organization in the story still contains classes, even though matter compilers make clothing, food, and covering available to all. I like to imagine how a just society would work.

The lack of artificial intelligence appears reasonable. The actual achievements of AI have always fallen short of predictions. Alan Turing estimated that by the year 2000 machines would be able to fool 30 percent of human judges during a 5-minute Turing Test. In 1965 H. A. Simon wrote that "machines will be capable, within twenty years, of doing any work a man can do." In 1967 Marvin Minsky wrote, "Within a generation ... the problem of creating 'artificial intelligence' will substantially be solved." Clearly none of these predictions have appeared.

I look forward to seeing how these areas--nanotechnology, scarcity, justice, and artificial intelligence--develop during my lifetime. Perhaps I will see some possibilities fulfilled.