The idea of artificial intelligence, its supreme capabilities and threats of turning humanity into dust has thrilled the minds of science-fiction writers and film directors over the past decades. Nowadays it becomes not only a matter of fiction, but science.
The pace of human progress multiplies with every year and opens the possibilities which were only dreams before. The efforts of industry giants like Google, Facebook and Baidu move us closer to the date when AI become a familiar and common thing like today’s smartphones.
OpenAI is the initiative of the world most significant executives – SpaceX and Tesla founder Elon Musk, LinkedIn founder Reid Hoffman, Y Combinator chairman Sam Altman and Palantir co-founder Peter Thiel - those who move the wheel of human progress today. The project was invested with more $1 billion over the recent years and meet no obstacles in its development despite the quarterly falls in the world economy.
The project’s goal is to create a beneficial AI that improves the humanity in a safe way – a machine comparable in its intellect with humans. However, most of the modern researches are now focused on the so called “deep learning” aspect of the AI.
Deep learning can also be called reinforcement learning as its process is a long caravan of trials and errors with goal of finding the right decision in various tasks. Another aspect is unsupervised learning when machines can learn data without being controlled and corrected by humans.
The project nearest agenda includes four technical goals. The first – is to define metrics allowing to measure the growth of intelligence.
The second – and probably the most practical one to date – is to create AI-basis for the production of physical robots that can perform household duties. There are a lot of robot-helper manufacturers even today but the founders’ idea is to give them AI that will make their robots do any task right off the shelf.
The third goal is to make a supreme chatbot, an interlocutor capable of integral and complex communication and comprehensive understanding of text – something that Siri and Cortana can only dream about.
The fourth goal – to build a single AI able to win a wide variety of games though there are already pre-AIs wining some simple ones.
The latter two points are likely happening in the nearest future: the latest NVIDIA project is basically a supercomputer developed especially for AI purposes. More than 3,000 people were working over three years to produce a machine with 170 teraflops of computing power – that is equal to 1 billion arithmetic calculations per second. Such capacity will allow AI to process an enormous amount of data with insane speed – enough to learn on its own.
The main problem of the AI? It can break free and wipe out the humanity of course! That topic has been outplayed many times in both cinema and literature but is there a way we can secure ourselves of such a threat?
Eventually, yes. Scientists all over the world recognize that AI can not only get mad but also get hacked thus it’s is extremely important to create an anti-AI system that will be able to secure, restrain and eliminate any possible threats. More over such a system should be created even before the AI itself to exclude even a minor chance of an AI-related emergencies.
As for the human factor – according to the Musk, “the best defense against the misuse of AI is to empower as many people as possible to have AI. If everyone has AI powers, then there’s not any one person or a small set of individuals who can have AI superpower.” Sounds smart, doesn’t it?
Taking into consideration the existing scientific and technological capacity of MENA-region it’s obvious that invention of such artificial intelligence will give a great boost to those industrial fields where one can apply this technology.
And while the world’s best minds are working on one of the most anticipating technologies it seems that AI future won’t be long in coming.