The age of Artificial Intelligence, popularly known as AI, is already here. It started off with being shown in movies and before we could blink our eyes, it took over the world of technology. Its employment in different areas is on the rise and so is the concern of the future that awaits us. This story is by Anuradha Gopal, an architecture student.

 

Narrator: Anuradha Gopal

Transcript

Good evening ladies and gentlemen, boys and girls, you are listening to Hello Educator.

(Music)

Today’s episode is from Anuradha Gopal and she is an architecture student. She’s generally very curious about a lot of things in and around her, starting from artificial intelligence to psychology to dance to music. So she has been doing a little bit of reading work on artificial intelligence and this story is on that. It just gives a little overview of what artificial intelligence is and how and where it is headed. This is definitely a good listen for all the little ones out there at your home. Happy listening.

01:02  The concept of AI. It’s something that I think we’re all familiar with, thanks to the numerous amount of TV shows, movies and novels. For me, personally, I find this to be a sort of a tired concept at this point in time. You’ve just seen it so many times in sci-fi and especially the idea that AI is going to take over like we’ve seen it in the famous movie Terminator. But, I’ve also seen this in some of my favourite works of fiction like A Space Odyssey or novels like I have no mouth and I must scream. Don’t get me wrong. I do find it to be a very interesting concept even though it’s overused but when Stephen Hawking in 2014 came out and said – “But I think that the development of full artificial intelligence could spell the end of the human race”.

01:51  The concept of AI taking over, it feels so far off because it seems so obscure the way it’s portrayed in Hollywood movies or in the work of fiction. But in reality, it’s actually what I think or what I have learned – it is an actual threat. But how did we come to this conclusion? Let me try and explain.

02:13  If we take a couple of steps back, in fact, a lot of steps back, there’s this game of tic-tac-toe that no matter what input you make the computer will never let you win. It’s kind of annoying. It’s programmed with algorithms so that no matter what move I make, it knows exactly what move to use to counter it so to make sure that I cannot win. That’s not that big of a deal.

02:38  In 1958, Herbert Simon and Allen Newell, they were AI experts and they said what if you could take a more complex game than tic-tac-toe? Say, just for example, there are a lot more possible outcomes in that game. They foresaw that within 10 years a digital computer would beat the world’s best chess champion. Now it didn’t take 10 years, not until 1997. You may have heard of this, it was quite a big deal at the time.

Deep Blue became the first computer that was able to beat the reigning chess champion at the time, Garry Kasparov. “On the blue’s 19th move, the champion resigns”.

Now it still doesn’t seem like that big of a deal and basically the way Deep Blue worked was that it would scan every single possible outcome it could make, about 200,000 per second that’s 2 lakh per second and it would make the best decision based on what he could find through this method of scanning.

03:39  At this point, I’m like – ‘Stephen Hawking, I’ve seen the videos of machines falling over ok? And I think we have nothing to worry about’. But here’s where I think it gets interesting. On 15 March 2016, the champion of the Chinese board game ‘Go’ was beaten by an AI called ‘AlphaGo’. This artificial intelligence was designed by Google’s DeepMind. It was a resounding loss. Now, the reason why this was such a big deal is that in Chess you only have so many options. But in ‘Go’, there are so many different moves that you can make. There are more possible moves that you can make than there are atoms in the universe, and there’s just no way that you’re going to be able to compute that amount of options to figure out what’s the best move to make. So how did they make this?

04:32  It may not seem like that big of a deal but it’s really cool. It basically uses deep reinforcement learning which is similar to how we learn as humans through trial and error, reward and punishment, and raw inputs. Say if we see something ourselves, the computer figures to learn itself. Basically, it uses neural networks to learn how to play the game which is similar to how we think as humans. And with enough computing power, you could simulate a human brain in this way. But we’re not there yet and it wasn’t this good from the beginning. It had to become good, it had to get good.

05:11  Not too long ago there was a viral video from a programmer by the name Seth Bling that uses a method to teach his computer to play Mario Kart. Mario Kart is a very popular video game. It didn’t know how to play the game but eventually it became good at it. In the beginning, it doesn’t even know where it has to go or what the options are or what even Mario is but eventually it figures out that it needs to move right and through different generations and learning from trial and error, adapting from these mistakes it eventually becomes better and better.

05:45  This similar method was used for AlphaGo, slowly becoming better and better and better and eventually becoming a master at this game. There’s another super cool video about a robot that does not know that it has limbs but it teaches itself how to walk despite this. So it’s just doing these random movements. It sort of figures out it has four limbs but it doesn’t know where these limbs are attached on its body and by trial and error it finally figures out where its limbs are positioned, and eventually it graciously moves across the screen. That’s really cool!

06:21  Self-learning AI is really cool and there’s a lot of advantages that we can use from this. There was a 3D printed cabin partition which was designed by a computer. It’s stronger than the original yet half the weight and it was used to fly in the Airbus 320, 2 years ago. So computers can now generate, come up with their own solutions to their well-defined problems. So then with Elon Musk as well as Stephen Hawking saying that AI could become a possible problem in the future. The idea starts to make a lot more sense to me now. They say ‘I think we should be more careful about artificial intelligence’ and they said that it could be one of the biggest existential threats that we have as humans. But the basic point which Elon Musk tries to make is that we have a general-purpose of learning the algorithm that evolution has endowed us with and it’s running at an extremely slow computer. Very limited memory size, ability to send data to other computers, we have to use this funny mouth thing over here. Whenever we have to build a new one, it starts over. It doesn’t know how to work.

07:32  So believe me, as soon as this algorithm takes experience and turns it into knowledge – which is so amazing that we haven’t done it in even software yet – as soon as you do know that, it’s not clear you’ll even know when you’re just at the human level. You’ll be at a superhuman level almost as soon as the algorithm is implanted in silicon. This was Bill Gates trying to compare how human brains and computers work. What he’s trying to say is that our method of evolving is very inefficient with comparing it with how AI would be evolving and exponentially growing and keeping in mind that humans are inferior without a doubt.

08:13  That being said, not everyone is on board with this idea that AI is going to take over or that it’s even going to be a problem in the future. AI can do a lot for us as humans. It can benefit us greatly, and I think what Elon tries to point out is that there are dangers involved with this development and we need to be careful. How can we protect ourselves from this? We are an intelligent adversary, we can anticipate threats and plan around them but so can a super-intelligent agent. For now, we can just use this technology and befriend this technology without letting it destroy us and let’s just let it for the future for us to find out whether AI is going to be a good thing or it could possibly destroy us in the future.

We sincerely hope that you really liked this episode. We upload one science story every day and if you are a science enthusiast and you would like to contribute to Mango Science Radio, please WhatsApp us at 9952243541 and you can also share your feedback via an audio message or Whatsapp. That would really be helpful. Thank you again, thank you so much for listening to Hello Educator. I will see you tomorrow.

Music by Karthikeyan KC