The film, 2001: A Space Odyssey might not be a total fiction anymore. Artificial intelligence is overtaking human intelligence at a pace higher than expected. Elon Musk, Bill Gates, Stephen Hawking are warning us over and over again that evolution in the field of AI is no less than summoning the Devil itself. It’s one risky job. Maybe, dealing with AI is tougher than dealing with North Korea.

Machine Learning(ML) is one of the buzz words that is used with AI, sometimes even interchangeably. But; ML is a type of AI that allows computer programs to learn when exposed to new data without being programmed. And guess who gives this “new data” ? Us. Humans. So effectively, AI learns ONLY what we teach it. For example, if you search ‘baby’ on you will only see white babies. Crazy right? AI is discriminatory because it learned that way. Maybe the concept of ‘black babies’ doesn’t exist for this machine.

This was a very subtle example of negative AI. Now, think of a world where there would be robotic weapons, or machines that can function entirely without guidance of humans (known as fully autonomous weapons). In a meeting in Davo, Switzerland there were talks of “killer robot” which tracks down and kills humans without human guidance. These robots could be of any shape and size. You won’t know if a person next to you is a killer robot, or if that dog you pet everyday is a killer robot. Could be anything/anyone.

AI learns by practicing over and over again. It’s right to say that- Practice makes AI perfect. This way, intelligence in machines will be substantially greater than the intelligence in humans. That means, AI can do ANY job a human does and only better. If you keep a machine on learning mode for a week, it will perform 20,000 years of human level intellectual work. And if this goes on, soon AI will make other AI and humans won’t be able to understand what’s happening. Once AI passes human level, it will accelerate exponentially. It’d be like an ant trying to understand the Internet. Machines are becoming more competent than us in every regard. What will be our purpose then? This makes me think about Charles Darwin’s Survival of the Fittest theory.

Facebook had shut down two AI programs after they discovered that they were writing their own code. Two programs even made their own language and started talking to each other. Nobody knows what were they talking about.

Sofia, first robot citizen of Saudi Arabia is known to have a dark sense of humour. She sometimes jokes about killing people.

AlphaGo is AI designed to play the game called ‘Go’. Go is a 3000 years old Chinese board game which is so complex that if you took all the computational power in the word and ran it for million years, even then the brute force wont be able to calculate the moves. And guess what! AlphaGo recently beat world champion in the series of playoffs. This machine approximated human intuition and outsmarted the best human brain in the game. This wasn’t expected to happen until next 10 years. But machines are already learning on their own.

Since last couple of decades, computational power is increasing. The power our phones have today is about a thousand times more than the power our PCs had 20 years ago. AI will only grow stronger and stronger as computational power increases because Google’s AI DeepMinds employs architecture of ‘Neural Networks’ that tries to mimics structure of human brain. The system uses nodes that act as artificial neurons connecting information. IBM’s Watson reads a lot of information and recognises patterns and builds up evidence to say Hey I am X% confident that this is the right solution to the question. All this requires immense amount of computational power.

The concept of AI was coined by John McCarthy back in 1955 but this kind of a revolution hadn’t taken place until now only because we didn’t have the computational power to do so. Imagine: IBM’s Deep Blue played chess and thought of millions of moves before the actual move. This couldn’t have been possible back in 1950s.

It’s a long shot to say that we are creating Satan or God. But we have great power. And with great power, comes great responsibility. Positive AI is our responsibility so that we are AI’s boss and not the other way round.

We always want to know what YOU think!

The WTF Files™



Leave a Reply

Your email will not be published. Name and Email fields are required.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>