Machine learning. Grandmoth!
Hey, Grandma, I didn’t meet you. But if you were alive, I would tell you that nowadays we teach machines as we teach children, only that they learn very quickly, there are many of them and they don’t forget. We teach them to imitate us, in our “intelligent” behaviors, such as driving, analyzing patterns, recognizing voices or winning a logical game, among other tasks that require cognitive ability, even if they only imitate us. But they are still somewhat clumsy, they are very good at specific tasks but lack multitasking and adaptation to diverse contexts. However, you can’t guess, they’re starting to learn for themselves, just like a child who says: Why? All through experience, like us.
Grandma, do you remember why you had so many children? Seven? Now, look at your daughters who on average had two, and what do you think of your granddaughters? Could it be that they have one son? As you can see, they have less and less, they have learned from your experience (data mining), right? they have extracted knowledge from your data, like your time, your circumstances, your life, they have really learned about it, they have advanced. And believe me, they are not only based on your experience, but on that of all the grandmothers they know, and not because your experience is bad, they need to cross more data to generate information they can classify, otherwise they will not have enough information to predict (Mathematics and statistics) and decide their future, so now they have more time to be, to progress, and especially not to depend. It’s incredible! Right? Grandma, that’s Machine Learning. Do you understand me? You’ve been smart, haven’t you?
That is what we are doing with machines, we are discovering the path that brings us closer to being truly intelligent, to being able to learn in context, to not being imitated, to being able to conclude while learning to learn.
And since I’ve caught your attention, what do you think if we take it to the present, as if you were here with me. I tell you that since you passed away, fabulous things have happened like the internet and computers, increasingly smaller and more powerful, which has allowed information to travel faster and anywhere, so fast grandma that she would see you first on the internet than in person. Imagine the same in just about anything you can think of.
There is so much data, but so much data, so many things but so many things that neither you nor I can handle in our lives, and perhaps not where you are now. But the surprising thing about our species is that it always has a creative way of solving problems that are beyond its reach (at least in a physical way and in terms of calculation capacity), and generating solutions with the creation of technology, a word that has gained a lot of relevance in the last three decades, because it has advanced exponentially since your departure.
You wonder why I exaggerate and make so much emphasis on the data and its processing, what I can tell you is that it is about the data that gestate the machine learning and therefore the artificial intelligence. We have reached a point where the capacity for storage, calculation, technology and applications have converged, allowing us to glimpse the potential that this concept has in the evolution of our species, as part of an extension of the capacity for human reasoning. But don’t panic like everyone else, I ask you to. It is only part of our evolution. We are creating a way to process patterns faster and more reliably, something we have always done, or you are going to tell me that you do not remember the patterns you saw in the grandfather, and especially because you chose it. You trusted him, you believed in a future together.
Well, that ability to see the future that you had with your grandfather, is what we are building with machine learning for machines, and we will probably achieve it, perhaps if we can make them intelligent. And the fact that we can give them the faculty to visualize the future does not mean grandmother that they determine it. Just like us, just like your daughters and granddaughters, they always depend on data.
It is in this dependence on data that mathematics takes on great relevance and gives meaning to the pattern that insinuates you from your grandfather, for nothing is more effective than mathematics in our understanding of reality and patterns. For it is in these, as a language of the known universe, that it is possible for now to describe reality and create others.
It is only natural that we are afraid that a machine can learn something for which it was not programmed, we have been the only ones until now, of course it is scary to be creating something in your image and likeness and that I failed you, what will God think of all this? Could it be that he had as much fear and doubts as we did? Or are we literally reinventing ourselves? I know you don’t know, grandmother, I don’t think any of us do, but this idea already lives with us, and that’s why I wanted you to understand it. Even from your cloud.
Since we understand the concept, which I have narrated in a figurative way from other scenarios and which I hope I have been able to transmit in such a way, that if you imagine the world of possibilities before us. Think of some, and let me insinuate one of those possibilities, probably subjective and obvious, but perhaps the most conclusive from my point of view, the possible replica of a human being.
Don’t look at me like that, grandmother, in my opinion, and based on literature and science fiction films, we have achieved each general concept to replicate a human being (at least in the only thing that differentiates us from other beings, the ability to learn), how would you say, according to my mother: “Time to time”. And so it is, make a fast and senseless parallel of the elements that allow our learning and machine learning: eyes to cameras (and so on with other senses and sensors), body to IOT (Internet of Things), brain to artificial intelligence, experience to data, and so on.
If you start to think and extrapolate our history, it seems that it will be like that, it is only a matter of time, of solving each question, of consolidating each concept, of building it, of having the resources. Really think about it.
Finally, and to let you rest in peace, wherever you are, I hope optimistically that you have more understanding of this space-time that I inhabit, and so, you could tell me if there is any possibility that we are a product of ourselves. I await your whatsapp.
References
- Understanding Machine Learning — https://futurism.com/images/understanding-machine-learning-infographic
- What is machine learning? — https://www.youtube.com/watch?v=WXHM_i-fgGo
- The Difference Between Artificial Intelligence, Machine Learning, and Deep Learning — https://medium.com/iotforall/the-difference-between-artificial-intelligence-machine-learning-and-deep-learning-3aa67bff5991
- Holberton Concept page: Artificial Intelligence, Machine Learning, and Deep Learning
