Machine learning works by looking at patterns, calculating the odds that one thing will follow the next.
Brains also look at patterns and predict how they go. It’s why we’ll often see things that aren’t there, or why we’re able to memorize things we chunk into patterns.
The difference between AI (artificial intelligence) and the human brain is that our brains do two things computers can’t yet do.
First, our brains consider context. Talking about shoe size can mean something different if you’re talking about a child or an adult, it can also mean something completely different in the context of a car.
In order to understand what we mean by “shoe size” we have to be given a certain context from which to identify patterns. Computers aren’t yet capable of understanding the complexities that surrounds context; things like who is asking the question or the environment it’s being asked in.
The other thing our brains do which computers can’t yet do is keep information connected outside of context.
If I say the word “size”, your brain can think not only of how it relates to shoes, but also: clothing, food servings, buildings or homes, animals, tv screens, phones, on and on the list goes.
Computers can’t yet connect one abstract or otherwise ambiguous term, like “shoe”, to all of the many, many different contexts it relates.
So when you and I have a conversation about coffee, for example, our brains are able to understand the context of a coffee shop as well as the notions of how the coffee shop often symbolizes community, conversation, air conditioning, wi-fi, tables, popular brands, business, and much more.
All the computer would know is linear information it’s been trained on. Machines are limited to the patterns we, as programmers, tell it to recognize. They can only learn language in so much as we tell them what a conversation might look like, then the machine calculates the odds of what a conversation entails as it occurs.
Tell an intelligent enough computer that you’re having a bad day and it might know to respond with “I’m sorry” but won’t have any concept of whether or not to ask you if it’s because of your boss, or the weather, or simply because it’s Monday.
It’s very likely we’ll have a computer developed in the coming years that can do each of the complex things the brain can do.
But for now, today, only you and I are capable in thinking this way; of being able to relate words and concepts not only to their contexts, but to nearly everything else.
That’s what makes creativity valuable: it’s vastly more complex than what even the most powerfully intelligent computer can do.
So let’s focus on doing that more often: understanding the context of what it is we’re working on, then branching out to everything else to see what connections there might be that a computer couldn’t catch.