AI Does Not Fucking Exist
An overly harsh headline to make a minor, but important point
In no place on earth does an Artificial Intelligence ( AI ) system exist. Nowhere.
To this point, no one, not big tech, not academia, not even a secret evil genius has anything approaching a platform that fits even the most basic definition of AI.
That definition, in simple terms; Artificial Intelligence is the automation of activities we attribute to human thinking and rationality, such as problem-solving, decision making and learning.
For most of us that definition is brought to life in applications like Jarvis from Iron Man, Data from Star Trek and most of the characters in Westworld and while no one is still certain we’ll get to that level of sentience, you get the picture.
So, I’m curious as to why it is we’re inundated with headline after headline, tech press and mainstream media alike, on how AI is omnipresent and taking over the world.
I haven’t been to any old west themed robot parks, have you?
This Company Is Betting a Little Artificial Intelligence Will Eliminate Your Bad-Hair Days — Inc Magazine
The Verge 2017 tech report card: Artificial intelligence and robotics — The Verge
5 Key Artificial Intelligence Predictions For 2018 — Forbes
AI Will Transform the Economy. But How Much, and How Soon? — NY Times
Welcome to the era of the AI coworker — Wired Magazine
AI is, I get it, a simple moniker, and it’s in the general area of what companies and governments are building. I’m probably not solving a significant issue with calling this out, but as the majority of the public has such little understanding of where the technology is at, it behooves us to be clear and inform accordingly.
What all of the headlines above and what almost every other headline you read that references AI is talking about is, in fact, Machine Learning ( i’m a fan of calling it Augmented Intelligence, though introducing an alternate use of AI would no doubt be problematic ).
While not nearly as sexy, Machine Learning is right now providing you immense value.
First, a definition; Machine Learning ( or just ML ) is the concept of using algorithms to identify patterns and / or make predictions based on a data set.
An example of this is Netflix taking all the information, from our viewing habits and correlating them in such a way as to predict that if you and I both watched a lot of show “A”, then you watched a lot of show “B”, it’s possible that I might also like show “B”, then displaying show “B” in my recommendations.
As data scientists everywhere cringe, let me be clear that the example above is a gross oversimplification of machine learning, but is essentially what a lot of those platforms are currently doing. As an example of that complexity, Apple uses Machine Learning to power FaceID in the iPhone X and collected over a billion images of faces to “teach” the system the patterns in how to identify people.
Other examples of Machine Learning in action: Siri and Alexa understand more of what we say, Tesla cars stop before they crash into things, Amazon recommends more products and Facebook suggests people we may know.
We’re using computational power and massive amounts of data to have machines make guesses, highly informed, calculated guesses, but not AI.
There will be amazing strides made in Machine Learning in 2018, bringing about a raft of announcements, but can we please call them what they are and not have everyone thinking that the artificially intelligent operating system from the movie Her, Samantha, is just around the corner.
As always, I’d love to hear if you disagree or can change my thinking, comment below.