Sadly, it’s still pretty accurate. Yes, we’ve come a long way [baby], but there are a number of factors that probably mean that this one will be with us always.
First of all, humans are still a lot better at certain things than computers. Now, a computer might be quite good at searching a huge database, or crunching numbers, or analyzing chess moves. Humans, though, are equally good at nuance, and vagueness, and emotion, and a different kind of complexity. Yes, with AI and machine learning, this is rapidly changing, but when it comes to the Turing Test, my money is still on the human.
Second, we need to remind ourselves that humans are still largely in charge. And what that means is that they still do the coding, and the design, and the requirements, and the QA – and basically put the darn things together. So, there’s still plenty of room for these error-prone humans to leave things out, to add in bad things, and to generally screw up the interaction between human and computer. It’s how our field got started after all. And – once again – I don’t see this going anyway anytime soon.
Finally, it seems to me that computers may also have finally passed a certain threshold. To me at least, they appear to be too complex for us mere humans to predict what they will do, how we should interact with them, what can go wrong, and how to fix them. Theoretically, we can get to the bottom of it all, but there are typically so many things in play, that it may take a lot of effort – perhaps even an infinite amount of effort – to really figure it all out.
There’s actually a book out there that speaks very specifically to this. It’s called Overcomplicated, and is by a Silicon-Valley-type named Samuel Arbesman. I haven’t gotten to it yet, but it’s #1 on my list. It promises to “offer a fresh, insightful field guide to living with complex technologies that defy human comprehension.” Can’t wait!
Love the sideburns!