_{Image: janjf93 / CC0}

Everyone agrees that human workers have been replaced by machines in the past. And everyone agrees that machines will continue to automate human jobs. It seems that technology advances rapidly nowadays and it is going to replace people from many, many sectors of the economy.

People agree on the above. But here is where they disagree. One camp (let's call them traditional economists and company) says that, yes, technology has taken people's jobs, but it always has simultaneously created new jobs and markets. So this time it will be the same - many jobs will go to the machines, but also many new jobs will be created and people will be employed to do them.

The other camp (let's call them information engineers and company) says that, yes, there will be new jobs created, but they will be very few. The vast majority of available jobs will be automated and consequently most people will be unemployed.

So who is right?

I don't want to go into statistics on how many jobs have gone from people to machines. I want to present my views on the topic because I haven't seen it laid out in quite this way.

I don't know the future, but a few things seem highly likely, based on current trends.

_{Image: Clker-Free-Vector-Images / CC0}

Let's first talk about the human brain. We can think of it as a general learning system. It can learn anything. If you live in an isolated tribe in a jungle, the brain will learn that environment, it will learn the language, the skills needed for survival there, all the cultural customs, etc., etc. Put a baby with the same brain in a modern technological society. It will learn the city environment, how to work with computers, it will learn about stars and galaxies, atoms, how to drive a car, etc. Our brain is capable of learning any sort of thing, including learning about itself.

Now, what about the machine "brain"? In 2012, a particularly interesting approach to building a machine brain became prominent. The approach is based on simulating neurons and how they connect to form neural networks. In the human brain, neurons connect with each other, creating biological neural networks. We can simulate this through computer software and thus build artificial neural networks. Just like the human brain, these artificial neural networks represent a general learning system. They can learn to do anything - from recognizing handwritten text or spoken speech, to translating between languages, to driving vehicles, to diagnosing diseases, to recognizing what's on images, you name it.

As an interesting sidenote, we are seeing more and more how the words "consciousness" and "intelligence" are now being separated. Recent examples are IBM's Watson which (or who?) won the Jeopardy game against the best players [1] and Google's AlphaGo which also won against the best players [2][3]. People don't consider these AIs to be conscious, but they do consider them intelligent.

So my point is that there doesn't seem to be a skill people have that machines will not be able to acquire. We seem to have found the approach - a general learning system - and it is only a matter of time before any task that humans can perform will also be able to be performed by an AI.

Even if there will be many new jobs created, it doesn't matter because the human capacity that allows people to do these jobs will be matched and outperformed by machines.

But can these artificial neural networks really learn to do anything?

I can't say for sure, because they haven't yet learned to do all the things people do, but so far they have been able to do amazing stuff. There isn't anything to suggest they won't be able to do the jobs of lawyers, managers, customer support, designers, writers, even the government.

Here are three examples. You be the judge.

A piece composed by an AI called Emily Howell:

The Tesla autopilot going for a test drive:

IBM's Watson vs the two top Jeopardy players:

So what do we do?

Two questions remain:

1. How long till great portions of the population are unemployed because of automation?
2. And, most importantly, how can we rearrange the social and economic system so all people can prosper in this highly technological society?

Recently I heard the famous chess player Garry Kasparov say in an interview that the future is a self-fulfilling prophecy. I think I agree with him - we will get whatever future we work towards.

References:
[1] https://www.theguardian.com/technology/2011/feb/17/ibm-computer-watson-wins-jeopardy
[2] https://en.wikipedia.org/wiki/AlphaGo_versus_Ke_Jie
[3] https://en.wikipedia.org/wiki/AlphaGo_versus_Lee_Sedol

---

_{If you like this post, follow me to get more}