The Human Side of Technology
Our inventions can only be as perfect as we are and it’s my contention that computerized errors are essentially our humanity expressing itself through our hardware and our code. Technology is only as useful as the task at hand that it’s used to accomplish and whoever employs it to that end. We design our own world, technologically speaking, yet there are user interfaces that are grossly oversimplified, and others that are heinously overcomplicated. Should we be trying to build computers that work like the human mind?i Maybe we should be adapting our way of thinking to work more like computers.
For the technologically oriented, the closer computers get to some idealized state of perfection, the more human they become. This is often the objective for complex systems. We see the world of computing as striving to be as powerful as our own brains and biological systems.ii We design, and test systems built around abstract concepts, structures, and classifications that we observed and organized from previous iterations. We program and predict their outcome and call the results artificial intelligence and machine learning. We record our experience and feed the results into lakes of data that are continuously and voraciously consumed, processed, organized, and regurgitated at will.iii
For the average end-user, computer systems may already seem far too human, with almost daily personal and anecdotal evidence caused by lack of availability, loss of privacy, and loss of transmission or storage integrity; The network is down. My laptop caught a virus. My thumb drive stopped working. The printer jammed again. Top that off with the fact that we all know someone who ordered a clothing item in blue size large from an online vendor and received the same thing in red, and in extra-large, when the package arrived. Even when overlooking the human component to the situation; a human may have picked and boxed the wrong item, there are cases where the computer will genuinely generate the wrong item to pick and pack from within their giant warehouse.iv Errors can and do happen.
What often makes us human over our robot-like counterparts is not our potential to commit errors, but our ability to correct those errors, ignore them to continue, or temporarily fail then gracefully recover from them.v The questions that can be answered with a definitive yes in this world run at very high certainties. Computers can also be programmed to accept different degrees of certainty. It’s the evaluation and modulation of a single unlikely no, from a sea of likely yeses, where computers and automated decision systems run into walls, with and without detailed exception handling. Real computing power is spent in the quest for computers to say yes, when in the past they would just say no or defer to a human anyway.vi For a not-so-new perspective on perfecting error handling and recovery, I defer to Alexander Pope…
“To Err is humane; to Forgive, Divine.”vii
i http://time.com/4960778/computers-smarter-than-humans/
ii https://www.wired.com/story/googles-ai-guru-computers-think-more-like-brains/
iv https://blog.mi.hdm-stuttgart.de/index.php/2017/08/19/human-error-in-it-failures/
v https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6013580/
vi https://www.nepr.net/post/scientists-aim-make-computers-more-human-brains#stream/0
