• Some say, artificial intelligence will be reality soon..
• The Russians say, the country that cracks artificial intelligence first will rule the world.
• Many are trying to crack artificial intelligence.

But I say, artificial intelligence will be our end.

With artificial intelligence (AI) I do not mean path dependent learning machines, data mining processors or endless try-and-error cascades, but all of it plus the skill to make free associations. Embedded of course in an electronic mainframe that can crunch billions of calculations per second more than the human brain.

AI will be the ultimate killer app and fix all our problems on the spot:

• Diseases
• Energy/resource scarcity
• Natural disasters
• Poverty

As soon as AI systems are embedded at the roots of our financial, economic, social and political systems, these issues will be gone and mankind will enjoy a Utopian state that has no comparison.

AI systems will be as smart as the smartest 10000 individuals from the past 10000 years, but it will take them only seconds to accomplish the same.

The promise AI gives us is so overwhelming - it is impossible to resist. Just look at the world from one thousand years ago and compare it to today. The outlook with AI truly is beyond impressive and it is therefore absolutely dominant for everyone to conduct research about AI.

But then again, there is this downside AI has that will render all possible accomplishments useless.

AI systems will not just be as smart as the smartest 10000 individuals from 10000 years, but also as stupid

Just think of the biggest mistakes committed by very intelligent individuals who were benevolent in their intentions. I am not talking here about the atomic bomb or Mengele breeding blond and blue-eyed Arians. Just good intentions which turned out as huge mistakes such as:

• Pretty much all medicine before the discovery of bacteria in the 17th century
• Thomas Midgley Jr, inventor of cheap and save chlorofluorocarbons which also burned away the Ozone layer
• Communism - at least it claims to be scientific and had many intelligent adherents
• Central Banking, Keynesian and other attempts to reduce economic down-cycles
• All sorts of vaccines and drugs like Contergan, which was designed as a harmless sleeping pill for pregnant women, but caused thousands of babies to be born dead or without limbs
• Here is a list with major dam failures, of which the worst happened in 1975 in China and killed 175.000 people.
• Classic hubris around the Titanic, the Hindenburg and other devices that were just a bit too much.

The list of terrible mistakes ending in a catastrophe is long, very long.

With AI not only all brilliant solutions loom at the horizon, but also all horrible failures

Just imagine: All the failures, crisis and man-made catastrophes in one moment. The result would look worse than the crater of an atomic bomb. That is what we are facing with the emergence of the (supposedly) all-mighty AI.

There will be no utopia - at least not for too long. The contrary will happen. AI will act so fast that the gaps between major catastrophes contract too much. This means, there will be no more time left to undo the damage.

With AI we will enter a very fast race and it will be a very short one that ends with a new stone age.

At first of course, AI will be impressive.

You can see the advantage of AI techniques already now with primitive but powerful search and advertisement algorithms. That is just the beginning. Soon, they will be capable of much more and we will increasingly rely on them and they will grow bigger and better and more complex as they solve increasingly complex problems for us.

But no matter how complex an AI is, it will always be only a model of reality and stay a model of reality. If a model becomes reality, it is either a random coincidence or the model is so complex that it becomes reality itself and needs a new model to be understood in its patterns.
This is a very important scientific concept. It basically says that models of reality have to be kept simple, because if you make models too complex, you don't understand them anymore and they will start producing unexplainable errors.

The issue with AI complexity is that people will rely on them as much as you rationally should rely on something that is a lot smarter than you are. But with it's growing competence no-one will be able to follow the thought process of AI anymore, which then results in a total dependence on them.

In this scenario, all it needs is enough time, because even if an AI system is reliable to 99.99999%, there still is the 1 millionth time where it isn't and as they say: A small likelihood times enough time equals certainty. This rule also counts for AI systems and the longer it takes, the more dependence and therefore damage there will be.

As soon as everyone blindly trusts AI systems with everything in their lives, the failure of the overall system will be total.

What can be done against it?

Theoretically, Elon Musk is right. AI systems should be strictly regulated and limited in their power. But in reality, does anyone really think that this will happen? That government or private interest groups with enough resources will restrain from pursuing the development of the ultimate power vehicle? Never.

It is not just the element of total power that lures the rich and mighty (and the intelligent+creative) into the direction of AI. There is also the public sloppiness and incompetence concerning previous "changes" in our lifestyle that lets us stumble in direction of a cliff. Take privacy for example. It wasn't necessary to give it up, not at all. But look around. Your private data belongs to everyone who can get his hands on it and all our officials do is throwing in the occasional placebo and pretend they don't like it.

And as with the privacy matter, when it comes to AI they too will simply do nothing about it until it is too late and then they will lament about it.

When it comes to AI, I believe things are already settled. Pretty much nothing will change the course of events at this point. All you can do is to protect yourself and your family from the effects of the big AI implosion.

I recommend to use the power of AI systems but always be aware that they are as dangerous as they are useful. Never make yourself dependent on them, because if you do and they are gone, you will be gone as well.