Do we think like computers? - Chapter 13

Ever since the development of digital computers, we have been told that our minds work like computers. Both are machines, both think, both can perform complex operations. However, on closer inspection the computer metaphor is misleading. Computers work in the world of clarity rather than the world of shadows. They perform best in well-ordered situations, with explicit knowledge/clear goals. They struggle with ambiguity.

Computers solve problems by establishing massive and comprehensive problem spaces and then finding ways to search those spaces. People solve problems by matching them to familiar patterns and spotting leverage points that have been successful in similar situations, forming connections based on previous experience, and working from there to construct a solution. We don’t search problem spaces.

How to connect the dots?

Chapter 12 explained why sensemaking is more than connecting the dots, but compares the ways computers and people connect a set of dots. A leading research paradigm (a thought pattern) for connecting the dots is the traveling salesman problem. The problem is to find the shortest path that passes through a set of points and returns to the origin. This task is analogous to a salesman traveling to different cities and eventually returning home. Such a salesman would want to find an efficient route so as not to waste too much time driving. The computer kept calculating and calculating to come up with shorter routes. After 1500 generations, the computer tour has approximated a reasonable solution similar to what a person would produce. So computers can learn and can improve.

Why would no human start with the random connections?

What if we double the number of cities? Humans would just double the amount of time they take to find a good solution. That’s because we don’t calculate the answer. Our minds don’t work like digital computers. There are many tasks that computers can perform better than we can. They can calculate more quickly and accurately. They can search more vigilantly and manage more complex situations. Nevertheless, there are some skills, that computers still struggle with.

Chess is very difficult to master. Computers approach chess through brute-force methods and lots of ‘crunching of different lines of play. Chess grandmasters rely on the patterns they have learned during many years of practice and playing to recognize good moves.

Why does playing chess requires more than finding the best moves in various situations?

For example, when Russian chess champion Garry Kasparov was beaten by the Deep Blue Program, this illustrated the human ability to diagnose problems. These things shows how critical and how difficult it is to make sense of feedback. Even though Kasparov lost, the incident further illustrates the magic of human cognition.

Ever since the development of digital computers, we have been told that our minds work like computers. Both are machines, both think, both can perform complex operations. However, on closer inspection the computer metaphor is misleading. Computers work in the world of clarity rather than the world of shadows. They perform best in well-ordered situations, with explicit knowledge/clear goals. They struggle with ambiguity.