In our previous article we explained about possibilities of modern Artificial Intelligence (AI) and promised to show you how it works. Are you interested? So, we are ready to move on.

Creating an AI, a person projected it to look and think as a human being. And what lies on the base of human intelligence? Right, neural network. And it is not that complicated as it seems, neural network – is basically just bunch of neurons that can communicate with each other. Almost as quantity of members in facebook, there are 10-20 billions of neurons in each person’s brain. Each neuron can save piece of information and make decisions, based on signals (or posts) from other neurons, and also “like” and “repost” information, transferring it further to other neurons. The more neurons we have – the more data our neural network can keep and process, therefore the network will be more intelligent.
rankbrain-artificial-intelligenceNeural networks are not programmed as we used to think (they are not based on strict algorithm), they are educated. Yes, possibility of education – is one of the main advantages of neural networks against traditional algorithms (e.g. ATM won’t call you and ask secret question for authorization, if you honestly tell him that you forgot your PIN-code, it only can act according to the programmed sequence of operations regardless the situation.

But let’s see how the neurons communicate in practice while building simple neural network. We will also explore what happens in mind of the person when s/he falls in love…


Okay, so let’s see how neurons work. Imagine that we have one neuron, let’s call him Bob. He knows how to analyze data from different sources and take decisions. When Bob meets a new girlfriend, in his mind he has such a condition:

1-  If I like her appearance and have what to talk about/ YES
0 – if I don’t like her/ NO.

And then a question arises: “In which kind of person Bob can fall in love with?”
What if the girl is beautiful, but not very intelligent, or maybe she’s got a Master’s in Mathematics, but she is not very good looking? So, based on which priority, Bob will make a decision with whom to fall in love?aiFor each of us one parameter (beauty or brain) will be more important. So, except the existence of the value of parameter, there is a level of its importance (or, as we call it, weight). So, if we multiply value of parameter (“yes=1” or “no=0”) by its weight (e.g. 80% from 100%), we will receive factors of “beauty influence” and “brain influence” on the feelings of Bob to a girl.ai2Simply saying, if we are giving greater weight to the parameter “brain” and smaller weight – to parameter “beauty”, Bob will fall in love with a clever girl. But if in total we don’t give enough value to parameters, Bob won’t fall in love at all. If the value will be accordingly 10% for beauty and 20% for brain, it’s still not enough to fall in love. Or, if we have more than 100% – then Bob will fall in love with anyone.neuron5

Basically, this is the main rule of how neuron works.

Artificial neuron – is such a function that reorganizes several input facts in one output fact, while doing this, we can also determine the weight (importance) of each of facts (there may be more than just “beauty” or “brain”), and also “excitation threshold” (or meaning of “falling in love”, in our case).

Therefore, the main outcomes we received about neurons in love and life are:

  1. We can educate neurons of Artificial Intelligence, and even make it fall in love, determining some specific set of parameters and its weights.
  2. If the weight of both parameters is too small – it won’t fall in love at all, or simply the parameter won’t work.
  3. If the weight of both parameters is too big – it will lose the ability to fall in love in according to specific factors (beauty/brain) and will fall in love in any kind of person.

In fact, for many people life science ends at this level, but now we know how to fall in love as neurons do, right?

Be Gera-Intelligent!

Sergey Slivkin
Vira Selentii