This article about AI and the philosophies behind adopting it can also apply to how we think about regulation of AI. Making a sausage may not be the prettiest thing in the world and maybe regulators aren’t going to be the best sausage makers the first time out (how long can I drag this metaphor out for the sake of one photo) but that doesn’t mean that we shouldn’t get started. If we are always looking for the perfect set of regulations that addresses every issue that can be thought of, nothing will ever be accomplished. Unfortunately, we are seeing very few items that can even be seen as precursors to legislative activity.
There are so many calls for things to be done ASAMATTEROFURGENCY which leaves the impression that unless things are handled perfectly, some impending disaster looms. There is really no evidence that this is the case. It is without dispute however, that “[t]he lack of clear regulatory and oversight mechanisms is concerning”. So let’s put some meat in the casing and turn the crank; we can always fix the spice mix in a second batch (yup, still going).
I would have added “seek the good, avoid the perfect” as one of my Collected Principles of AI Regulation but it’s not really a principle that applies specifically to AI regulation.