The time has come to tackle my hardest article yet. In this article, I will try to explain to you, very simply, what Artificial Intelligence consists of, why it’s so important, and what it will mean for your future.
In my last 2 articles, I talked about machine learning, and deep learning – all small subsets of the beast that is AI. Feel free to read them, and tell me what you think, because this next one will be a big beast.
I will now attempt to describe to you the 5 most important pillars of this technology. Lets get started, shall we?
1. Artificial intelligence makes a machine THINK
Thinking. All of our lives are spent doing it, even while we sleep. Despite that, have you ever thought what it is that makes you a thinking organism? What even is thinking?
Thinking is defined by many as “using your mind to reason about something”. And it just so happens that your mind has a lot of tools at its’ disposal to do its’ purpose.
Here are some of the main tools:
Now ask yourself – do you think a computer can do any of these things?
The answer, to those who read the article on Machine Learning, is obvious: no. A computer like the ones we use today simply takes what you put in, does what you want it to, if it’s programmed to do it, and spits out the result. That’s it. It learns nothing, does not wonder why you wanted the task done, does not plan for the future, doesn’t even communicate with you! The messages you see that ask you a question have been made by other humans and put into the computer.
Artificial intelligence, as you may have guessed by now, fills the gaps in these weaknesses. Generally speaking, it consists of giving a machine some or all of these qualities normally associated with humans, and therefore ally the computer’s main advantage (doing arithmetic and processing algorithms much faster than humans), with human qualities, making a tremendous powerhouse in the process.
Now you know the main concept. But how can we humans give a machine intelligence?
2. There are many different ways to create an AI
Lets go back a bit. Remember how I said a PC only does what it’s programmed to do? Well, how does it do this, you may be wondering?
I may cover programming in a later article, but for now, lemme drop the magic word that is the bread and butter of a computer’s working - algorithms. An algorithm is a step by step instruction of what a computer should do. When you tap a key on your keyboard, when you click on something, the PC follows algorithms that tell it what to do in that situation. And strangely enough, the key to have PCs do more than what algorithms allow them to, is to use more algorithms.
There are algorithms that, against expectations, tell a machine to go above and beyond to learn from it’s environment, and go beyond it’s programming. More importantly, there are many different kinds of algorithms that can do this, and I’ll briefly explore them now.
• Symbolic AI
This type of AI focuses on one basic premise: that the human mind is built on symbols, that humans can read, and that by forming connection between these symbols, a program can achieve reasoning. This type of AI would simply be fed a ton of information related to its’ purpose, and it would work from there on in processing the info.
Many AI systems have been developed with this paradigm, such as ones built to use human logic methods to solve problems, or ones with knowledge banks stored within, which they can access to solve problems.
• Non-symbolic AI
This type of AI focuses on another entirely different paradigm – machines can never learn to imitate a human mind, so why bother? Simply develop mechanisms for the AI to receive information and learn things on its own, without relying on pre-packaged info.
Types of systems that had this paradigm in mind are, for example, machine learning. Fuzzy logic systems (I may discuss these in a later article), and other approaches that focus on computation.
All this technical data must be putting another question into your mind though. What exactly is it that these systems can do. The short answer?
3. They can be made to do ANYTHING
Whereas Machine and Deep Learning had certain purposes in mind, an AI (which encompasses them both) can be made to do literally any intellectual task. It has gotten to the point where you can’t go by one week without seeing at least 3 articles discussing something at least vaguely related to AI.
From healthcare to data mining, to self-driving cars, to games, to search engines, there is nothing you cant make an AI do. Gonna bulletpoint the most important examples.
AI can help doctors and hospitals in many ways. For starters, it can help doctors diagnose many diseases, simply by inputting the systems, making up for many common human errors. Secondly, as the medical world leans more and more to Big Data, it becomes increasingly more important to have a system that can interpret that data and extract the valuable information. AI can do all of that, with an ease that astounds the doctors that get to experience it!
Have you ever driven a self-driving car? Guess what powers them. There is an efficient AI behind these systems, and it has a great track-record, with many studies pointing out that these cars can drive safer than humans. Another unexpected thing that AI can do better than a human operator!
AI promises to solve one of the biggest threats of nowadays – hacking. These systems can find out which files are safe to download in the blink of an eye, as well as stopping phishing attacks, and much more powerful cyber-threats. They do this by analyzing many different variables within a computer (IP address, network traffic,…) and point out when something is wrong.
It’s worth pointing out that hacker attacks powered by AI could be more devastating as well, but stronger protective systems might just be more worth it overall!
I’ve mentioned potential applications of Deep Learning and Machine Learning systems is previous articles, and they mostly describe the scenario. A well built AI can perform tasks like predict stock prices, and weed out any fraud attempts much better than any human so far. They are already in use for these ends, with a good track record, and are predicted to only get better as time goes on!
All of these applications are pretty good, right? Every rose has its' thorn though. Time to find out the problems surrounding this technology…
4. They pose many problems
As with any technology big enough to change the world, this one has it’s own suite of problems attached. Ai theorists and researchers themselves are somewhat scared of the impact these systems could have if not held in check.
“The development of artificial intelligence could spell the end for the human race”. Words uttered by Stephen Hawking himself. The famous scientist believes creating a technology that can at some point take charge of itself and evolve to a point where we cannot match them could be a disaster, should the AI change it’s priorities. Not a Skynet level disaster, but it’s worrying nonetheless.
Others raise economic issues – Ais promise to make a lot of jobs obsolete. How would we compensate the workers who would lose their jobs on account of this? The recent tech advancements are now raising new issues such as universal income.
Another issue, more on the social and ethical side, is what this technology means for us as a species. The ultimate goal of artificial intelligent, is basically to create an intelligent being. Do we humans have a right to do this? And what status would these new beings have, if we achieve them?
There are many divisive sides on these issues. All the more reason you should look this up on your own, and make your own judgements on the situation. Because one thing is for sure.
5. They are the future – for better or worse
Computer science has many things going for us. In one of my first articles, I spoke of quantum computers, and the revolution they offer. There are always many new different things going on in the tech world.
Despite that, as I follow news on Reddit, and many other sources, I find that many of the articles coming up are from authors discovering new application and uses for AI. And the numbers agree with me. Artificial intelligence research occupies 10% of all computer science research going on. For such a vast field, this is pretty huge.
There are already many AI systems existing and growing in the world. Do a Google Search. Likely, an AI is helping to narrow down the results to show you exactly what you want to see on the first page.
Did you buy something on Amazon, and get a recommendation that was exactly what you wanted to buy next? An AI did that too. Not to mention Alexa, the AI system they sell that can do pretty much anything you ask it too (except kill your boss hehe), that you could do on your PC.
Like cars? You’ve definitely heard of Tesla. They’re leading the field in AI research to make their self-driving, absolutely cutting edge cars.
What do you think Siri is? Like Alexa, they are AIs in constant growth, and in some years, you may have a conversation with them like you would with a normal person, such is the progress going on in this field.
Your children will likely grow up in a world where AI is not a sci-fi dream, but an everyday reality. If that doesn’t blow your mind, I don’t know what will.
That about wraps it up!
This seems like the right time to end this article. It’s the biggest I’ve ever done, and I hope I got all the right info, without boring you!
I don’t have much of an idea what to write next. Feel free to sound off in the comments below, I always like talking to you guys :)
If you have any suggestions in how I can improve, or some criticism, or some questions, feel free to let me know!