mathNEWS Issue 110.3: Friday, June 12, 2009

It's okay to assume

One thing that bothers me is when people have problems with assumptions. Hilariously, they assume that all assumptions are bad, which is an assumption in itself. Assumptions are a way for us to simplify things, especially when we notice a trend. Which brings me to my comment: it is not insensitive to assume that people are heterosexual. Let's assume that I meet a male and have a conversation with him. Then I ask "Do you have a girlfriend?", I argue that it is not insensitive to ask this question.

Assuming that the male I meet is actually homosexual, I would expect some response similar to "I'm gay and have a boyfriend" or just "No" since that would answer the question. By saying anything other than "No" means that the male has assumed that I wanted to know anything other than whether or not he current has a girlfriend. It also doesn't assume that he's a polygamist in which case the answer is also "No" since he doesn't have a girlfriend, but he has several girlfriends.

Now, the most generous estimates of non-hetersexuals that I've seen is 12% of the human population. Assume that this is the actual value. Now assume you that have a file which will work on 88% of all computers. If you see a random computer, you will assume that your file will work on that computer since the majority of computers are compatible with your file. It's only logical. You can accept that when the file doesn't work, that this is one of the computers which it doesn't work on, but the expectation is that the file is going to work.

In articial intelligence, many methods involve a form of participation for classification. In these cases, predicting future cases is done by taking a majority poll of some form. If the majority is heterosexual then guess what? The prediction will be heterosexual.

Everyone makes assumptions. I find most people say that assumptions are bad. I think assumptions are a great tool to help simplify complex decision making. We make assumptions all the time, but there are only certain ones which are considered acceptable. The trend that I've noticed is if it is an assumption which contradicts something a politically correct zealot would enforce then it's the bad kind of assumption.

Angelo



Copyright © 1998 mathNEWS.