Hashtag Artificial Intelligence
The internet is awash with stories about something called Artificial Intelligence. Confusion around what it is is prompting many to proffer definitions of it, or corrections of wrong definitions given by others. Here’s one that was presented recently at a talk by one of Amazon’s top Machine Learning people:
Artificial Intelligence: A system or service which can perform tasks that usually require human intelligence
This is a fairly common way to define it. Here’s a similar formulation from Nathan Benaich in his post 6 areas of AI and machine learning to watch closely:
The ultimate goal of AI […] is to build machines capable of performing tasks and cognitive functions that are otherwise only within the scope of human intelligence.
One problem with this definition is that it means the state of being an instance of Artificial Intelligence is temporary. If we create a system that is capable of performing some task that usually requires human intelligence, then once it is widely adopted it is no longer true that the task in question usually requires human intelligence. This leads to what’s known as the “AI effect”, which I wrote about before.
Another problem is that it excludes tasks that can be performed by non-human animals, such as locomotion or object recognition.
But slippery or otherwise problematic definitions don’t stop the term popping up all over the place. “Self-taught artificial intelligence beats doctors at predicting heart attacks,” reads the headline of a recent article about applying basic statistical techniques like logistic regression and random forest to medical data. “This AI learned to predict the future by watching loads of TV,” reads another equally nonsensical one. Every day we’re seeing more and more headlines like these in the mainstream media.
What is Artificial Intelligence? It’s a meme; an impressively resilient and fecund meme. No sooner does it land in the brain of one unsuspecting human than countless new #ArtificialIntelligence tokens are spawned and sent out into the world. So much power does this meme wield that people doing good work in various fields are being forced to define what they’re doing in relation to it. Just look at all the “What is the difference between Artificial Intelligence and Machine Learning?” articles out there. Really, why should Machine Learning be defined in relation to AI?
The slippery definition issue above can be looked at as follows: it is the term “Artificial Intelligence” looking for things to refer to. It latched onto things like Optical Character Recognition for a while. Right now it is joyfully riding the coattails of Deep Learning.
Once we’ve exploited Deep Learning to its limits, #ArtificialIntelligence might hibernate for the winter (yes, we may well see another one of those), but will eventually latch onto something else. Always along for the ride will be the “pernicious fantasy,” to borrow a phrase from Daniel Dennett, that machines with human-like intelligence are just around the corner. This is no more true now than it was 50 years ago but many smart people are utterly convinced of it.
The term “Artificial Intelligence” has been around since the early days of computer science, when “thinking machines” were seen as the natural next step after programming basic logic. Decades later, we have a better understanding of the complexities of human cognition, informed by evolutionary biology among other fields, and we see that things like the internet and Facebook were more natural progressions from formalized computation than were thinking machines. Machines continue to be tools for solving human problems (they don’t have problems of their own to solve!). But the idea that this somehow constitutes intelligence on the part of the machine lives on.
Don’t panic though — I think the #ArtificalIntelligence meme is mostly harmless, apart from sowing confusion and perhaps a little fear. But it’s important to be aware that just as the “chicken is only an egg’s way for making another egg,” each one of us may only be #ArtificialIntelligence’s way for making more tokens of #ArtificialIntelligence.