The Inmates are Running the Asylum is an evangelical calling for interaction design in software development. Currently, there is much discussion regarding interaction design in information architecture circles. Cooper, who is known as an interaction designer, gives information architects a wonderful non-technical introduction to what the current situation is, where his firm is going, and why it's important to do good interaction design.

Alan Cooper argues that interaction design methods can help us develop more powerful yet simpler software products that users will get more pleasure using. To do this he asserts that interactive products need to be designed by interaction designers instead of by software engineers.

Starting off with basic ideas about computers and how they have changed products and how products are used, he then goes through a litany of examples and common mistakes, such as "dancing bears," made in software product development. The book finishes by giving us three methods of his "Goal-directed design" methodology, which we can use to start incorporating interaction design in software development projects. Cooper's three methods are designing for pleasure using personas, designing for power using goals, and designing for people using scenarios.

Cooper is trying to solve the problem of bad software design, or software that makes people feel like they are idiots. He also rails against the current mentality in which the user is constantly apologizing that it must be their fault instead of the infallible, perfect computer. Cooper lays the blame of users' difficulties firmly on bad software design.

To me he trounces on the usability "straw man," saying it has been co-opted into the bug-testing phase of development. In the world according to Cooper, only "DESIGN" (done before anyone starts programming a line of code) will save the day. He wisely notes that although management BELIEVES programmers can do interaction design as a part of programming (as well as information architecture), this is a wild fantasy. Unfortunately, the result is bad interaction design (or information architecture) that is designed solely to please the programmers who wrote it, and not at all pleasing to the people who actually use the software.

In general, there were too many times I felt his arguments were forwarded by zealous opinions rather than scientific results or at least measured accounts. I think his heart is in the right place, but his logical path is paved over with so much ethos that it chokes me even though I do really want to swallow his argument.

If you are wondering what battles interaction designers are currently fighting, this is a great introduction. I recommend paying attention to his war stories so that hopefully we can avoid the same mistakes and protect software development from the characters he warns us of, such as the "computer apologist" or "jock programmer."

My favorite part of the book was Cooper's list of users' "personal goals" that software should not violate (pp. 154-6). The product interaction should not fail any of these four tests.

• Not feel stupid.
• Not make mistakes.
• Get an adequate amount of work done.
• Have fun (or at least not be too bored).