There's no denying of the impact of "Artificial Intelligence" - what took 50 people to do in the past can be now done by 5, entrepreneurs are finding it easier and quicker to scale, and retrenchments and re-skilling of the labour force are going to be the new norm.
But what really is "Artificial Intelligence"? The term seems to imply that human intelligence is replicated/re-created through artificial means, in this case through advanced statistical models, a massive amount of data, and investments the size of a small country's GDP.
And that implication sets us off thinking it's only a matter of time before Terminator/[insert your favourite dystopian AI movie here] becomes a reality, and humans can go extinct (humans have many many ways of going extinct, it's a miracle we are still here).
Don't get me wrong, I do understand AI is an incredibly powerful tool/weapon that can cause catastrophic damage in the wrong hands (or even in the so called right hands), and ethics and governance in building AI needs a lot of attention.
However, I'd like to pause for a moment and reflect. What really is "intelligence"? And can that "intelligence" really be replicated by statistical models & computing power?
Intelligence - what is it? A simple google search gives this: "The ability to acquire and apply knowledge and skills."
The ability to acquire knowledge and skills - in the domain of biological life, many sentient beings do this in one way or another. A human baby learns rapidly by watching and mimicking her parents & other people around her.
A crow's chick learns rapidly the knowledge and skills relevant to a crow's life in the same way. And many others in the fauna world do so as well. And some have the ability to acquire knowledge and skills much faster than others.
Apply knowledge and skills - again yes, this is done by many sentient beings, in the way that is relevant for them.
But there is another underlying aspect that is not explicitly spelled out in the simple google search - the choice to learn, the choice to make mistakes, the choice to apply the knowledge in the way that suits them best.
And this is where human intelligence (and the intelligence some other animals/our pets too) is superior, because we have the choice to acquire the type of knowledge we want & apply them in the way we choose.
Yes there are societal limitations, but let's ignore that for now.
What does choice mean? A human can choose to lead a life that he/she wants. Every day, a human makes so many choices. And one way to look at it, is that a person is where he/she is due to the combined effect of all choices made up to the current point in life.
We have the choice to make mistakes. As many mistakes as we want, for as long as we want.
Not every choice is a good one, not every one is a bad one. Some choices which appear to be bad, puts a person on a good path eventually. And when do we call someone "intelligent"? It is when we can visibly see the demonstration of intelligence, either academically, musically, artistically, through their craft/work, socially, or through their character. All through their choices made in life.
Now, what is "Artificial Intelligence"? It is an advanced statistical model, (statistics is basically maths but driven by data), trained using data to provide the best answers possible. "Best" as defined by the creator of the model.
In other words, it is just a system of nuts, bolts, gears - a big machine where the parts all link together using mathematical methods. It is trained to give the right answer by feeding it a HUGE amount of information, on what is the right answer/best answer to a question.
You can imagine a factory with a huge production machine, where data is fed in on one end in a funnel, and the machine chugs, and processes it through its many compartments and gives out a nice package at the other end. Do we call this machine "intelligent"?
But we do see creativity from "Artificial Intelligence", how is that? Well, it is driven by randomness. To allow for creativity and variety in the answer, some randomness is put into the machine. The architects of the machine find the perfectly right amount of randomness that gives the machine its so called "quirky personality". It's basically like throwing dice and modifying the output based on the outcome of the throw.
That is not creativity, it is just randomness.
Now what does this really mean? "Artificial Intelligence" is no doubt powerful, but it's definitely not a substitute for human intelligence. And most importantly, we should be careful not to allow AI to dumb us down, not dull the sword of our own intelligence.
Rather, we need to use AI as a tool, to better acquire knowledge and skills that we choose to. To better apply the knowledge and skills as we see fit. To help us make better choices...or better mistakes, whichever we want.
The role of "Artificial Intelligence", is to augment, or enhance human intelligence.
Never to be a substitute of human intelligence.
ความคิดเห็น