Ok, you see the headline sounds like a blast, thinking I'm going to read something academic right? Actually my intention at the beginning with this article is to write it very academically. But I found the casual style is much more interesting, let's start. The whole content of the article will address the root of wisdom and what is changing the evolution of mankind today.

Once upon a time, long ago, there was a great philosopher named Aristotle, he spend all his life seeking the right answer to the question, "What is the purpose of human?" It is written for easy to understand (actually, every philosopher wonders that question). Aristotle offers a philosophical doctrine called Teleology, which states that all things in nature exist for the ultimate goal (the Greek term for telos). Things exist for the sake of other things, and things exist for the sake of perfecting the essence. Although that view is somewhat theological, it is the foundation of many later philosophical theories, typically Hegel and Marx.

Teleology has much influence on Biology in the study of evolution, to answer the question "Why did you come to this Earth?" Sorry I'm joking a bit (too many headaches here: D). But really, the question of the origin of human intelligence originated here. Why in the same habitat where homo sapiens become so intelligent in a different way? Why not a sabre-toothed tiger? Nor the dolphin? In 1842 Charles Darwin penned the first sketch for a book called The Origin of Species. In that Darwin gave us a bit of light about the origins of human intelligence, the answer is natural selection. From the 8th grade, you knew "natural selection", right? What natural selection has to do with human wisdom?

800 thousand years ago, the human is still much like a chimpanzee, not as beautiful as it is now, Link (a homo, I am playing Legend of Zelda so ..., well) are hunting in pouring rain, suddenly he saw a wild pig running in front of him. Link is in ready to prepare slanting poor piggy, then suddenly the sky blast! The wild pig caught fire and burned in front of the Link, "So lost this one today," Link thought and turned sadly prepared to leave. Suddenly, a strangely scented wonder never caught the Link's nose, he found strange and turned to look at the burning fire. Link approached and stared at, then tasted, "Oh my god, it's awesome!" Well, the story I just told is the origins of rotten pigs, oops sorry, the origins of fire: D Fire can be considered as one of the greatest inventions, and it is actually found so (in an understandable way). In the lengthy flow of natural selection that lasts millions of years, mankind is often one step ahead of any other species when it comes to finding things like weapons, fire, and later wheels, steam engines, electricity, the Internet, etc. And those inventions are the source of human intelligence.

So why mankind dominates the world but not another species, it's a question that not only Mario Balotelli wondered, "Why always me?": D In 1998, in a prominent statement "Variability selection in hominid evolution", Dr. Rick Potts hypothesizes why we are, not another species. The answer of a few hundred pages, but can be summed up in one sentence, is the ability to adapt extraordinarily to the external circumstances of human beings is the key to make human becoming as smart today.

By joining Doraemon's time machine and fast-forward to 2019, we live in an era where adaptation reaches the upper limit. People do not want to do boring and repeatable work, they need to create something in replacement. And that's when people "summons a demon" (Elon Musk) that we called, artificial intelligence, or AI (Artificial Intelligence).

Well, AI is now a trendy word, everyone is trying to add AI to whatever they are doing. I've seen companies that only add AI-based features to their old products and make them into fashionable brands and can sell for as much as five or ten times. But this article does not mention what negative is happening in the AI ​​field. In the narrowness of the content, I just want to mention AI as an evolution of humanity. And indeed it is so, whether speaking literally or figuratively.

AI, or Artificial Intelligence, is a field whose father is now widely recognized as Alan Turing, one of the world's first computer scientists. The famous Enigma machine created during World War II was one of the first forms to help Turing have the original idea of ​​artificial intelligence. Eighty years later, the AI ​​has grown extraordinarily, and all the top scientists and thinkers like Stephan Hawking and Elon Musk are worried about Singularity. Did you watch the movie Terminator? Skynet (at the moment, Chinese AI-based surveillance system called Skynet, how creepy!) is an conspiracy version of the Singularity, when machines begin to think intelligently as humans. So, the question is, could AI be so intelligent, could replace human, become the most intelligent species in the world? The answer is yes, but subject to the following conditions:

First is when the AI ​​begins to simulate the organization state of people. Among the greatest inventions that change the mankind, organizational forms are one of the main causes of human's overtake of other species and mastery of the world. The way people organize tribes and then, form the government is what has completely changed the order of nature. The wisdom of mankind has some basic characteristics, one is accumulation, the other is inheritance. Accumulation means that human beings 500 thousand years ago are not as intelligent as they are now, human intelligence has been buoyed by tens of thousands of years, plus a number of great leaps as mentioned above that make us today having a huge amount of accumulative knowledge. Second is inheritance, the main accumulation over tens of thousand years of changes in the genetic structure and neuron structure of the human brain, is due to natural selection, so people can inherit the super gene from the previous generation. Genetics has made people increasingly complete, imagine you put a newborn baby in a pack of monkeys (like in Tarzan), the child will grow up with the intelligence of a monkey, but when you bring it back to the human world, its intelligence immediately adapts and grows to the human level. We can not do the opposite, put a monkey into the human world and hope it will be as intelligent as humans. But AI can, and that is the second condition:

The second condition is that human succeeds in simulating the function of the brain and reproduces it in machines, in a simple way that man creates an artificial brain and inserts it into the head of a brain's machine. AI scientists have succeeded in simulating the function of the senses. For example, we now have Siri, which can handle natural language (though not very intelligent), which is the simulation of listening and speaking. We now have image recognition tools with 99% accuracy, AI now can distinguish who this face is, this block is the TV or the barrel, which is the simulation of the vision function. . We have DeepMind who won the human in Go, that is the simulation of thinking function. Fortunately (?), we have not created AGI (Artificial General Intelligence). In fact, AGI is the holy grail of all AI researchers. We now have two main research areas on AGI and in fact this article is based on my thoughts after reading OpenAI's latest AGI research.

In the scope of this article with the main goal of non-technical discussion, so I will not go deep into the technique. Alright, the first direction called Gradient Based Optimization and the second direction is Random Optimization. The first, or more commonly known, is Deep Reinforcement Learning, which is the direction that Facebook and Google are leading. Especially Google with DeepMind super team, the next challenge of DeepMind machine is the human victory in the game Star Craft. If Deep Mind did, then Google has taken a huge step in creating the AGI. The second, or genetic algorithm, or Black-box optimization, is a subdivision of evolutionary algorithms. This algorithm simulates the principle of natural selection to gradually produce better, more understandable results, this approach makes AI as in the evolution of human intelligence (which I have explained above). This trend is being supported by Elan Musk with OpenAI and called Evolution Strategies. On March 24, 1974, OpenAI in a statement entitled "Evolution Strategies as a Scalable Alternative to Reinforcement Learning" shook the Machine Learning / AI industry with results from their research. And, surprisingly, this initiative is not too confusing and easier to use. Again, people are more anxious (or happier?) with the imminent release of AGI.

So that, for me, as an AI / ML researcher, AI is the next evolution of humanity, its importance is like inventing fire or the Internet. In order to end this blog, I want to give my thoughts, that AGI will be created but not in the other two research directions. Looking back at the great moments of mankind, fire was not created from research, but from human observation of natural phenomena, Charles Darwin proposed the theory of evolution based on philosophy, Albert Einstein found Relativity wherever I couldn't imagine :D Do you see, whether the theory of evolution or the theory of relativity is the foundation of biology or physics later but all are still theory - means people who has not proved Darwin or Einstein to be right or wrong. And I think the creation of AGI is the same, it will appear in a random moment, it just happens to be like the evolution of humanity. And is that the end or the beginning of a new chapter of mankind? Let's see :)


P / S: I am looking for more colleagues to study AI / Machine Learning with very specific and practical applications to hundreds of millions of users and changing fundamentally global internet infrastructure, to whom may be excited about, please email kevin@uiza.io.