Growing up I always thought when AI was achieved it would be something like a robot or computer that was almost sentient. Obviously we haven’t reached those levels yet but I still constantly hear about all these things that are being done with AI. So in simple terms how does our current AI work?
In: 0
There’s A. LOT. of current things that could be considered AI. Any specific AI you’re curious about how it works?
Current AI works on machine learning: algorithms that let you find a good solution for the task witout even knowing how this solution will work like. Just by tweaking some parameters. A repost of my older comment on the same topic:
So, imagine playing a game: you are told a number, you add some X to that number and tell the result. You will be told if the result differs from the one expected by the person who told you the number, so you have to guess the correct X.
“1. What do we want as a result?”
“Well, maybe X = 0? 1+0=1, my answer is 1.”
“No, for 1 we need something bigger. Let’s try again, what do we want to get for 2?”
“Then maybe X = 2? 2+2=4, my answer is 4.”
“No, we need less than that. Another try: what do we want for 3?”
“So, X is bigger than 0 but smaller than 2… Maybe X = 1? 3+1=4, my answer is 4.”
“Yes, that’s what we needed, you guessed the correct X!”
In this scenario, “take a number and add X to it” is your *algorithm* and X is a *parameter* for that algorithm. You don’t know that parameter beforehand, you guess it in an iterative way only from the required answer.
Turns out, we can construct an algorithm with quite a lot of parameters (possibly, millions) in such a way that there *will be possible values* for that parameters which, in theory, will give us good results for the task at hand. Not perfect, but good. We don’t know *what exactly* these values are, we only know that they *can exist*. The task can even be as complex as showing the algorithm an image of a bird and expecting the answer “bird”, it still may work with some parameters unknown to us.
Learning methods allow the program, in a similar way to the example above, start with a completely random guess and then tweak all these parameters in a more or less sensible way only based on what the expected answer is. And the math goes in such a way that it will likely slowly find better and better combinations until it encounters something that actually works to an extent. This process is what’s called machine learning and the set of values found for the parameters is called a *model* for this specific algorithm.
AI is a *very* broad umbrella. Really, it just refers to *any* piece of software that’s designed to “intelligently” tackle a problem, but that “intelligently” is open to a lot of interpretation. For instance, you can play the game of tic-tac-toe comparatively intelligently, and comparatively unintelligently, but playing intelligently is actually unbelievably easy. Writing an AI to play tic-tac-toe is trivial. But all the same, as a human observer, you’d be able to see a clear difference between a trivial AI, and a program that just randomly picks squares. There’s something in there that we recognize as “intelligence”.
Of course on the other end of the spectrum, you can write a program that let’s a car drive itself. Not so trivial. But all the same, you’d be able to tell the difference between a car driving intelligently, and a car swerving around at random. The particular problem is much harder than tic-tac-toe, and the AI vastly more complicated, but it’s the same sort of subjective judgement being made.
So how then does AI work? Well it depends. You could write a tic-tac-toe AI in which you’ve literally hardcoded every possible board state along with the ideal move to make in that situation. It’s probably not something you’d *want* to do, but you could. Alternatively, you could write an algorithm that plays out different variations internally, and attempts to find a move with which it can force a win no matter what the opponent does. Or you could use machine learning (itself a whole *host* of different techniques) to train a model that can recognize which board positions are winning and which are losing. Ultimately, the approach you take is going to depend on the problem you’re trying to solve. For turn based abstract strategy games like tic-tac-toe, or chess, or checkers, you’ll use something called minimax-search. If you want to automatically label photographs, you might train a neural network. If you want to schedule meetings to minimize conflicts, maybe you’d go with something called integer programming.
So there isn’t really an easy answer. Something is AI if we subjectively judge it to be behaving intelligently, and how the AI actually *works* is going to depend on the problem and the people who wrote it.
Intelligence, down to its core, is just a combination of knowledge database, decisions, and logic processing.
AI, is basically a program or code that allows machines to “emulate” the decision-making skill of a human with basic knowledge and logic. For something to be considered AI, it needs to be able to gain info, understand said info, then respond to info.
Lets take Siri for example: when you ask “hows the weather”, Siri “thinks” about what you are asking by looking at its database, then “understands” the context. It then decides on the next action to take, which is “look up the weather”, and finally gives you the answer.
A program or code that does some of these but not all is not a “ai” but algorithm.
A program that is designed to program itself and/or other programs, to achieve a specific goal.
For example, a site like Amazon will use AI algorithms to show you the most relevant products to what a specific customer would want to buy, based on your data compared with millions of people’s data. If customers aren’t buying a specific product that’s shown to them, the AI will tweak itself to show that product to different people, and the algorithm’s success rate goes up if it starts showing that product to people who will more likely buy it.
Or an AI program could play through a level of Super Mario World 1,000 times, making slight tweaks to each button press and each move, until it successfully completes a level as quickly as possible.
Latest Answers