Don Norman's rambling The Design of Future Things is just as thought provoking as his earlier books, but three times as succinct (I am looking at you Design of Everyday Things) . In it he discusses what he sees as the future of intelligent devices, namely autonomy and human ability augmentation.
I found his discussion of "loose reign" vs "tight reign" control to be one of the most interesting parts of the book. It is an excellent metaphor that could be usefully applied with some sort of literal interpretation. My thought, going to his focus on the intelligent car, would be to have the steering wheel be able to be pulled out or pushed into the dashboard a certain amount, where fully extended would be maximum human control and pushed in could mean more car control (perhaps folding out of sight completely if the car is fully automated). It would be interesting to see what other things this could be applied to, where an automated system continues navigating about (perhaps the web?) until a human grabs hold and shows interest.
RANDOM TANGENT!!
One thing that continued to pop up into my mind while reading this was why noone has tried putting neural networks or the like into videogames to test automated learning techniques. Perhaps people have, but I think it would be fascinating to integrate AI learning into an online game, where an AI could watch players techniques, whether in strategy games or FPS or racing games, learn from them and then try them out itself. It seems that if you wanted to test how effective an AI system is you could first test it for basic design flaws within a virtual environment populated by lots of human users and test it there.
No comments:
Post a Comment