I’m from the old school. I suppose, that sentence alone makes me seem like a codger. Let’s call it the eighties. Part of the art of problem solving was to work toward a solution and get it as tight as we possibly could before we committed to implementation. It was called the design process and today it’s called “design thinking.” So it was heresy to me when I found myself, some years ago now, in a high-tech corporation where this was a doctrine ignored. I recall a top-secret, new product meeting in which the owner and chief technology officer said, “We’re going to make some mistakes on this, so let’s hurry up and make them.” He was not speaking about iterative design, which is part and parcel of the design process, he was talking about going to market with the product and letting the users illuminate what we should fix. Of course, the product was safe and met all the legal standards, but it was far from polished. The idea was that mass consumer trial-by-fire would provide us with an exponentially higher data return than if we tested all the possible permutations in a lab at headquarters. He was, apparently, ahead of his time.
In a recent FastCo article on Facebook’s race to be the leader in AI, author Daniel Terdiman cites some of Mark Zuckerberg’s mantras: “‘Move fast and break things,’ or ‘Done is better than perfect.’” We can debate this philosophically or maybe even ethically, but it is clearly today’s standard procedure for new technologies, new science and the incessant race to be first. Here is a quote from that article:
“Artificial intelligence has become a vital part of scaling Facebook. It’s already being used to recognize the faces of your friends in photographs, and curate your newsfeed. DeepText, an engine for reading text that was unveiled last week, can understand “with near-human accuracy” the content in thousands of posts per second, in more than 20 different languages. Soon, the text will be translated into a dozen different languages, automatically. Facebook is working on recognizing your voice and identifying people inside of videos so that you can fast forward to the moment when your friend walks into view.”
The story goes on to say that Facebook, though it is pouring tons of money into AI, is behind the curve, having begun only three or so years ago. Aside from the fact that FB’s accomplishments seem fairly impressive (at least to me), people like Google and Microsoft are apparently way ahead. In the case of Microsoft, the effort began more than twenty years ago.
Today, the hurry up is accelerated by open sourcing. Wikipedia explains the benefits of open sourcing as:
“The open-source model, or collaborative development from multiple independent sources, generates an increasingly more diverse scope of design perspective than any one company is capable of developing and sustaining long term.”
The idea behind open sourcing is that the mistakes will happen even faster along with the advancements. It is becoming the de facto approach to breakthrough technologies. If fast is the primary, maybe even the only goal, it is a smart strategy. Or is it a touch short sighted? As we know, not everyone who can play with the code that a company has given them has that company’s best interests in mind. As for the best interests of society, I’m not sure those are even on the list.
To examine our motivations and the ripples that emanate from them, of course, is my mission with design fiction and speculative futures. Whether we like it or not, a by-product of technological development—aside from utopia—is human behavior. There are repercussions from the things we make and the systems that evolve from them. When your mantra is “Move fast and break things,” that’s what you’ll get. But there is certainly no time the move-fast loop to consider the repercussions of your actions, or the unexpected consequences. Consequences will appear all by themselves.
The technologists tell us that when we reach the holy grail of AI (whatever that is), we will be better people and solve the world’s most challenging problems. But in reality, it’s not that simple. With the nuances of AI, there are potential problems, or mistakes, that could be difficult to fix; new predicaments that humans might not be able to solve and AI may not be inclined to resolve on our behalf.
In the rush to make mistakes, how grave will they be? And, who is responsible?