Thinking about the future, we either dream about how cars will carry us anywhere we need to go about their daily business and help us make the perfect solutions. Or scared of the massive displacement of people from the economy soulless robots. But what goals really should stand before the interaction of man and computer? The publishing house “Mann, Ivanov and Ferber” published the book Microsoft CEO Satya Nadella “to Refresh the page. On the transformation of Microsoft and the technologies of the future in the first person”.
- When one artificial intelligence talking to others
- To come to the Board through the avatar and print on the printer a frog heart
- What do we do with artificial intelligence, which has its own plans
- 4 scenarios for the future that are worse than the TV series “Black mirror”
- The machine can’t sit down and say, “Hmm, what should I work?”
As should ultimately evaluate AI (artificial intelligence) — how useful or damaging to humankind phenomenon? I’m inclined to believe at first. But that is exactly how it happened, first we need to go beyond thinking in the spirit of “machines against people”.
Unfortunately, too often science fiction writers and pioneers in the field of technology drawn into the game with contrasting electronic human mind in a kind of war for supremacy.
AI Deep Blue, IBM was called in 1996, a sensation, proving that the computer is able to win over a human chess match of the championship level. A year later, Deep Blue took a giant step forward by defeating legendary Russian chess player Garry Kasparov in a tournament of six games. It was a stunning sight — the computer took over the human chess game, has long been regarded as the pinnacle of the human mind.
In 2011 IBM’s Watson won two masters of the TV show Jeopardy! (the TV quiz game, popular in many countries around the world. The Russian equivalent is the quiz show “jeopardy”). And in 2016 AlphaGo Google DeepMind beat Whether Sedalia, the South Korean master of go is an ancient complex strategy game played with pebbles on a Board the size, usually 19 by 19 lines.
Without a doubt this is the tremendous achievements of science and engineering. But the future in store for us even greater surprises than the victory of computers over people in different games. As a result, people and machines will cooperate and not oppose each other. Just imagine what will open up opportunities when man and machine will work together, struggling with the greatest problems of humanity: disease, poverty, illiteracy.
But that AI has approached this level, the efforts are much more ambitious than launching rockets to the moon. The head of our research lab in Cambridge Christopher Bishop once made a Memorandum, in which he argued that we need to put as much effort as it requires a whole programme of space research — many parallel, single but related rocket launches.
Our goal is to formulate a great and inspiring challenge, which is of great public importance, which needs AI.
Venture capital investments and transactions in this area are on the rise, but their main goal remains unclear.
In 1961, President John F. Kennedy said that America is the great goal before the end of the decade to land on the moon. Such an objective was possible mostly due to the associated Grand challenge in the technical field, but also because of the need in global cooperation.
Similarly, we must formulate a target for the AI, which will be quite bold and ambitious. It must go beyond just what can be achieved through gradual improvements of existing technologies. The time has come for universal cooperation in the field of artificial intelligence.
Steps in this direction have already been taken. In 2016, without much fanfare, Microsoft, Amazon, Google, Facebook and IBM announced the creation of a partnership in the field of AI for the good of all people and society as a whole. The goal is to help the community better understand the AI problem, and to develop better methods of solving problems and exploiting opportunities in this field.
Established partnership will advance research in the field of development and testing safe AI systems in such areas as automotive, health care, cooperation between man and machine, economic imbalance and social good.
I got some idea about what can give significant public developments in the field of AI, watching the performance engineer Microsoft Sahib the Sheikh, who participated in the development of payment technologies of vision — Sahib who lost it in early childhood. Taking advantage of advanced technologies, including pattern recognition and advanced machine learning, Sahib and his colleagues have created software applications for the small computer that can be worn as sunglasses.
This technology is able to resolve the ambiguity and interpret data in real-time, shaping the picture of the world and passing it to the Sahib in audio form instead of visual. This tool gives the Sahib a richer perception of the world: for example, it is the street noise with the tricks performed by the skater, and sudden silence during a meeting — so, what’s the participants. Sahib may even read a restaurant menu because the computer whispering in his ear the names of the dishes. And, perhaps most importantly, he is able to find in the crowded Park, his friends gathered for a picnic.
In numerous discussions about the future of AI overlook the great opportunity of collaboration of machines and humans. It seems that our perception of the AI was stuck somewhere in the middle between the Intrusive voice of the computer-killers HAL from “Space Odyssey 2001” and friendly votes of a modern personal digital assistants: Cortana, Siri, and Alexa.
We either dream about how to dispose of suddenly fallen on us free time when cars will drive us where we need to go about their daily business and help us make the perfect solutions. Or scared of the massive displacement of people from the economy soulless robots. It seems we are not yet able to go beyond this choice between utopia and dystopia.
I bet that is where productive will be of no use in discussions about AI opposition “good — evil”, and identify values, followed by people and companies engaged in the development of this technology.
In his book “Homo Roboticus? People and cars in search of understanding,” John Markoff writes: “the Best way to answer difficult questions about control in a world full of intelligent machines, is to understand what qualities have those who build them”. An interesting observation that our industry should adopt.
At our developer conferences, I explain Microsoft’s approach to the problems of AI based on our three core principles.
First, we want to create artificial intelligence that enhance and complement human capabilities and experience. Instead of thinking in the paradigm of “man against the machine”, we intend to focus on how human characteristics, such as creativity, empathy, emotion, sensitivity and intuition can be combined with the computing power of artificial intelligence, i.e. the ability to analyze large amounts of data and faster to perform pattern recognition in the name of social progress.
Second, we need directly to build trust in our technology. It is necessary that technology is initially able to protect the privacy, transparency and security. Devices based on AI should detect new threats as they evolve to create new ways of protection.
And thirdly, we have created technology should be available to all, they must respect the needs of each person, bringing people beyond cultures, races, nationalities, economic status, age, gender, physical and mental abilities, and so on.