The third law of prediction from the late great Arthur C. Clarke, is that “any sufficiently advanced technology is indistinguishable from magic.”
If Sir Arthur were writing today I think he may have replaced ‘magic’ with ‘artificial intelligence’.
AI has become our modern sorcery. It’s both our savior and our bogey-man. It will most likely show us how to be better human beings right before it destroys our worthless pathetic lives, because it’s a vindictive demon unleashed by godless computer scientists.
We define AI, like magic, as anything we don’t completely understand.
Is Google artificially intelligent? If not, why not? You ask it questions, or just type in a few words, and it goes out and returns information from all over the world related to question you asked or the words you entered.
“Well,” you say, “it’s a just complex algorithm running on really powerful servers, that takes into account the words I searched for and their prevalence on certain web pages, and then it returns those pages in order of decreasing popularity. That’s not really intelligent.”
Most of us do not consider Google ‘artificially intelligent’ because we understand it. Or more accurately, because we have a general heuristic to explain how it does what it does.
Watson winning Jeopardy? OMG, it’s AI! Oh, well actually, the computer was fed the clues in an electronic text format while the other contestants read/listened to them. Then it searched it’s massive data banks for relevant answers and gave a response. Kind of like the “I’m Feeling Lucky” button on Google’s home page, if it provided answers in the form of a question. (What are ‘Pictures of Pamela Anderson?) Incredible, amazing, but if you’re like me somehow less impressive when you know it wasn’t listening or optically reading the questions. Why? It doesn’t really diminish the accomplishment, but it also feels somehow less “intelligent”. I think the problem is that, like the Wizard of Oz, we’re a little less impressed once we glimpse the man behind the curtain.
As an alternative to Clarke’s law, let me submit: “the more you understand how a technology works, the less likely you are to think it is magic/AI.”
So when Altman Weil asks this question about AI, it’s no wonder they get the responses they did.
- In 2011, a MAGIC computing system called Watson defeated two former Jeopardy! game show champions, demonstrating the power of MAGIC. Since then, Watson’s performance has improved by 2.4 BA-JILLION percent and the IBM MAGIC Group is reportedly working with a number of legal organizations on a variety of MAGICAL applications for the profession. Can you envision a law-focused MAGIC ‘Watson’ replacing any of the following timekeepers in your firm in the next 5 to 10 years?