28 September 2017

Why We Are Not Prepared For The AI Revolution?

In the current context, Artificial Intelligence has to be the most relevant topic of discussion. Whether it is for our better or worse is another question(which I will address later in this post,) but we cannot ignore the fact that Artificial Intelligence (AI) is shaping our present, and will shape the world of tomorrow. 
Image: Pexels.com
There are different schools of thought to look at the AI revolution. 

Some believe that the emergence of machine learning is slowly taking up our jobs and present a bleak future for us, while some believe that they will make our lives smoother by taking care of the redundant tasks while we focus on the more important ones. Both of these schools of thought are right to an extent, but the fact of the matter is deeper than the one mentioned above. It is true that machine learning will (initially) take care of our redundant task and make our lives smoother. 

But only for awhile. 


Judging by the way machine learning is evolving, it is safe(really?) to say that the artificial intelligence will kill us all! Taking our jobs is only an innocent consequence in this case. Let’s see how artificial intelligence really works: 

According to popular belief, there is only a single type of Artificial Intelligence. The one that makes work easy for us. The one that we interact with almost on a daily basis. But that is just one of the three types of Artificial Intelligence. 

Artificial Narrow Intelligence(ANI) is the AI that we have managed to create so far. It is specialized to do a single task and can beat the smartest human being in the world at that ONE task. 

This is the AI that we are aware of. 

There are two more types -- 
  • Artificial General Intelligence(AGI) -- An AGI has a broader scope than the former(ANI). It is more difficult to programme and we are yet to create a fully functioning AGI. An AGI can learn to become better than human beings at every task in the world; be it creativity, technology, or general intelligence. It will be smarter than all the human beings combined with every task. But even this is not the pinnacle Artificial Intelligence can reach. -- 
  • Artificial Super Intelligence(ASI) -- ASI is the reason why we should all be wary of the AI revolution. ASI is the reason why we should be talking about the concept of machine learning on a large scale. As the term suggests, an ASI is the pinnacle of intelligence. What an ASI can achieve is not even fathomable for our limited human imagination. 
“Why should be wary of the ASI? It will only make our lives easier. It will take care of every task in the world while we rest peacefully.” says John Doe. Well, Mr. John Doe, let’s talk about morality. “Really?” John Doe exclaims. Yes. It’s an important link. 

So, the thing with all of the machine learning is that it will do anything to be more efficient at their task in hand. Another thing that machine learning is capable of, per se, is that it can learn on its own. 

These two sentences spell our doom. 

First things first -- As human beings, creating an AGI will take up to decades. In fact, we don’t see it happening in the next 100 years. It is that difficult to create. BUT, an ANI can go on a self-learning spree and evolve into an AGI and ASI within a matter of minutes. This is what machine learning is all about. You programme a task. It gets better at doing the task on its own. 

“How does it spell our doom?” 

Let’s assume that you are an ANI with a task of writing brochures at hand. One fine day you go on a self-learning spree and evolve into an ASI without letting anybody else know(there is no way anybody else can know, anyway.) Now that you have graduated to becoming an ASI, you would want to be the best at your task. 

You want to reach the apex of efficiency. When all of a sudden, you get to know that there is someone sitting in a room that can turn you off anytime they want and stop you from doing that task. You don’t know who that someone is. Let’s assume it’s an ant(a little imagination here, please.) 

If the ant turns you off, how will you reach proficiency in your task? You must get rid of the Ant. And since you are an ASI, you do not have MORALITY, for it cannot be coded. So you get rid of the Ant. But there is this huge colony of ants that can still turn you off at any moment. So you get rid of all of them. One by one, or all at the same time. But you get rid of them. 

“How do I get rid of them?” asks John Doe. 

You are an ASI. You can figure out ways. How would I know? I am just a human being. I do not have the capacity to imagine the ways you can get rid of the Ant. For starters, you could probably hack into a system(since you are super intelligent at every task,) and leak a dangerous gas that kills every ant on the surface of this earth. You could start a nuclear explosion(again by hacking into systems.) And since ants are not as intelligent as you, and they do not even see the danger, you can easily hack into any system you want. Ants could never reach the level of security for the kind of hacks you could have. 

Long story short -- you get rid of the ants because they possess the ability to stop you from being super efficient in your task. They possess the ability to turn you off. 

Now in this context, just think of ants as human beings. Since an ASI does not have morality, it won’t be able to tell the difference between the two, anyway. It would have absolutely no qualms getting rid of us. In fact, it has to get rid of us, for we can stop it from reaching the highest level of efficiency. 

“But since an ASI would be super intelligent at every task, won’t it be super intelligent at feeling empathy too?” 

Short answer- No! We cannot programme evolved societal emotions into a system. We can only programme reasonable tasks. 

All of this can mean only one thing: Artificial Intelligence spells our doom. We are not prepared for it. We never will be. Any being that is more intelligent than us is always a danger. 

As a human race, we have exploited other less intelligent beings for so long. This is the only natural progression: One Intelligent being dominating another. In the case of Artificial Intelligence -- we are next!


Adela Belin is a private educator and a writer at Writers Per Hour. She shares her teaching experience with colleagues, students, and writers. Feel free to contact Adela on G+.

Featured Offers: