With this post, I will wander a bit out of the Programmers at Work box and take a little interlude to digress to current events.
I went to hear Dan Ariely, a professor at MIT who works both with the economics department and the Media Lab, discuss his research and book Predictably Irrational. He was an entertaining, insightful speaker. His work focuses in on how people behave in decision=making trying to codify and explain seemingly “irrational” behavior. One example of the research he did was on cheating and how he found that there is a low-level of cheating that is widely practiced and tolerated in our society, yet over a certain threshold or in certain conditions, for example if you sign a pledge to be honest before you fill out your IRS form, rather than after you fill it out, you are much more likely to be honest or not. It was a fascinating talk and I’d urge you to go see him or read his book.
One thing he mentioned is the desire to use this research for “good” and to inspire a new category of software he dubbed in the category of “mind” programs designed for the consumer, the idea being to help regular people, individuals navigate their personal path through to a point of making a good, informed rational decision or plan in important areas such as “retirement planning”, health insurance, and the like instead of “irrational” ones which is more often the case. I loved this idea and thought it’s a rich territory for developers today, particularly those who are willing to focus on the “consumers” needs and foibles, rather than being a tool of the businesses who sell these services.
I left the meeting wondering if there was a counterpart to Behavioral Economics in the software field, perhaps called “Behavioral Programs” or “Behavioral Systems and in searching on the web I have found very little specifically called this. Yet, one thing we know today is that software could use a lot more understanding and appreciation of patterns of “irrational” behavior and the internet provides a vast connected knowledge-base and connected consumer base and through this much prediction and behavior can be studied and used. I suppose the work of start-ups such as Farecast and others are just the beginning of treading in this interesting territory.
I’m sure there are lots of others. So I welcome your insights and ideas about this area of programming and research.