Blog: All Your Questions About AI Answered
AI (Artificial Intelligence) is the buzzword nowadays, and many people are keen to explore what it is, with many questions about this seemingly new technology. This article will answer the commonly asked questions about AI and hopefully give the readers a better sense about what AI really is.
I keep seeing AI pop up everywhere, so what exactly is it and why is it so popular now?
Artificial Intelligence, like its name suggests, is an area of computer science that works towards making computers become as intelligent as humans.
Contrary to popular belief, AI is not a new concept — it has been around for a while — but it is only now that AI has managed to ‘take off’ because of the improvements in computing power and technology.
You see, in the past, there were only mostly ideas and theories in the field AI and little practical advancements because computers back then were simply not powerful enough. Also, to make a good AI, you need data (lots and lots of data), and it was only after the turn of the century that massive amounts of data started to get generated with the advancement of computing and communications technologies.
So now with the combination of increased computing power and data, AI practitioners are able to make really good AI models that are able to make significant impacts in the different industries and aspects of life.
Wait, so how does it impact me, this one little insignificant human living in Singapore?
Well, to put it simply, AI is starting to be everywhere now. There’s AI in the phone you use, the websites you visit, the car you drive, there’s AI everywhere! In most, if not all, industry sectors, AI can be incorporated to improve aspects of it, be it increasing the efficiency of production lines or allocating resources and manpower properly to the parts needed the most. Whether you like it or not, AI will probably impact you one way or another, maybe even without you knowing it.
People keep saying AI will take our jobs away and we will all be jobless. Is my future a bleak bottomless hole if I’m not doing something related to AI?
The sentiment that some new technology will come and ruin jobs forever is not a new one. During the Industrial Revolution, many workers felt the same way when machines came and made a lot of the processes that were previously done by hand automated. But what came out of the Industrial Revolution was not massive unemployment that lasted for decades. Instead, new jobs and roles emerged which made use of the machines to improve and make workers even more productive.
Such a thing will mostly likely happen too in the years to come. Some existing jobs will cease to exist, and evolve into newer roles. Retraining will be required for workers that previously worked in such jobs and as long as workers are able to accept and adapt, there would not be major issues. The governments around the world are also aware of such a change in the workforce and will respond to it with appropriate policies, with the progressive ones already doing so.
Sometimes right, I hear people say Machine Learning instead of AI, so what is the difference?
AI is sort of the all-encompassing umbrella term in which there are more specific areas of study such as Machine Learning. If you have consumed enough mainstream media in the past few years, you definitely would have heard of these terms — Machine Learning, Deep Learning, Neural Networks, Reinforcement Learning. Let’s look at what these terms actually mean:
Machine Learning: A collection of techniques where computers are trained to perform well at specific tasks without the need to hardcode any specific instructions into them
Deep Learning: A subset of machine learning that uses neural networks to train computers. This is currently the popular and the better performing area of AI.
Neural Network: A concept in Machine Learning that was originally inspired by the neurons in the brain, although neural networks have since evolved to be much more than that. The study of different types of neural networks and how to improve them is what Deep Learning is all about.
Then then my mother’s friend’s cousin’s child is apparently going to this course called Data Science and Artificial Intelligence, are those related?
Like I mentioned a few questions earlier, to train good AI models, we need lots of data, good data. You can already see the relation there.
Data science is a field which uses various tools, algorithms, mathematical and statistical methods to extract information from data and AI kind of does the same too, but in a more automated manner. So AI and Data Science share many commonalities and that’s why we often see the two fields of study intertwined.
How about the fears about machines taking over the world? Very scary leh!
The current ‘AI’ is nothing close to that, don’t worry. What we have now is narrow AI, which can be very good at doing specific tasks only, because they are optimised to learn for specific tasks. What ‘robots taking over the world’ would require is for them to achieve sentience and that’s a long way off.
You know hor, I saw the other day that the robot beat the champions in DOTA, is that one AI also?
Yes! That’s an area of AI known as reinforcement learning. It’s where machine learning scientists train the AI to maximise rewards in a particular situation. In the case you mentioned, the scientists from OpenAI trained their bots to win the DOTA match, to put it simply. What’s impressive is that the OpenAI team had five bots that were operating independently of one another and they had to learn to cooperate like an actual human team would.
It’s also the area of AI where many other headlines have been made, like when DeepMind’s AlphaGo beat the world’s best Go players. Go was previously thought to be a game where an AI couldn’t possibly learn to be the best as it involved many possibilities and strategies.
Wah like that AI seems very amazing leh, I want to try AI also! Is there anywhere I can pick up some skills?
There’s many online courses out there, but for most of them, you will need some semblance of programming experience and involves quite a bit of mathematics.
If you’re someone who has no prior programming knowledge or have forgotten a lot about programming but want to get start on both programming and AI, you could join our June Holiday Camp: AI Explorer. It’ll involve us teaching Python from the basics up and then ease you into the world of Artificial Intelligence. You’ll do a lot of hands on projects to not only understand how AI algorithms work, but to also create projects you can show off to your friends, like an image classifier (think Pokedex, or something that tells you what breed a dog is) or making an AI that can tell how happy your ‘Tweets’ are!