MAGAZINE CIS • # 1 (7) • 2019
Now you and I very often hear the words “artificial intelligence”. Someone to the question: “What is it?” Answers: “A, program.” Someone brags about their phone and how he knows how to happily answer questions, and some even compete with whom the program in the phone is smarter, and for someone the “smart phone” is an everyday tool.
The convenience of using a “smart phone” as an interface for working with the Internet, working with contacts and searching for information using voice is no longer a novelty for anyone. But, probably, not everyone knows what is hidden behind these opportunities, not many people understand what is really hiding behind the words “artificial intelligence”.
Over the past few years, many IT giants have been working to create and develop their versions of artificial intelligence systems. Many IT manufacturers no longer give their programs just names, such as Google Assistant or Google AutoML , but names, such as IBM Watson , Apple Siri , Amazon Alexa , Sharethrough Hemingway , Microsoft Cortrana and others, where the name of the developer’s company appears at the beginning of the name. …
I personally like one of them – Yandex Alice.
Let’s start from the very beginning and in order.
That is to say, from simple to interesting!
What exactly is “artificial intelligence” (AI)?
Yes, we all watched science fiction films and read science fiction, and we more or less imagine what it is.
But, it is not so simple to answer this question unequivocally.
AI is not just a program that speaks with a human voice. This is a large scientific area that is being formed at the intersection of such scientific disciplines as microelectronics and robotics, nano and biotechnology, medicine and psychology, computer science and information technology, philosophy and many, many others. The result of the activity, which should be the creation of a kind of perfect humanistic intellectual system, capable of creativity and self-reproduction.
Nevertheless, to any of us, AI seems to be some kind of perfect program capable of almost independently, either from its own database (knowledge base), or from the Internet, or from the outside world, to collect, accumulate and analyze huge arrays of structured and unstructured information, and then on its basis, make certain simple or complex decisions, presenting them to us in one form or another.
We are already dealing with this process everywhere. For example, AI through our smartphones with you, which we use in you as an interface for communicating with the program, provides us with background information and answers our everyday questions, for example, with the sweet voice of Alice.
Or more complicated, today there is an AI system that analyzes the condition of a sick patient, makes diagnoses to him and, based on the current situation, makes a decision about an operation that saves his life, while also “monitoring” the state of a person during the operation and at the stage of his recovery. Virtually excluding any possibility of death or any mistake.
Today the scope of AI is huge. It is used in industry, medicine, science, education and, of course, in games and entertainment.
In my opinion, one of the most interesting, promising and necessary applications of AI in the near future is associated with the task of searching for extraterrestrial civilizations and exoplanets suitable for human life, where it is necessary to perform complex calculations and analyze simply gigantic amounts of data collected by satellites, telescopes and various systems.
Another interesting trend in the use of AI is the direction of “smart city” and “city of the future”.
When was artificial intelligence (AI) invented?
The question of the date of the appearance of the term or of the artificial intelligence itself remains interesting.
A certain collegiate councilor, Semyon Nikolaevich Korsakov (1787-1853), studied the possibility of strengthening the human mind with the help of automated machines. Description of his inventions, the so-called “intelligent machines”, Korsakov published in 1832 in the so-called period of the first industrial revolution, during the dawn of steam engines.
Nevertheless, the first appearance of the term “artificial intelligence” in 1956 is attributed to the outstanding American scientist and programmer John McCarthy (1927-2011).
Active work on the study of the problems of artificial intelligence in Russia (then back in the USSR) began in the 1960s.
We are living in an amazing time, in the so-called period of the fourth industrial revolution. Literally at the very beginning. You can even say at the point of explosion of science fiction and engineering thought, during the period of the emergence of dozens of new technologies, and most importantly, during the period when it became possible to put artificial intelligence into practice.
One of the innovators in this area in Russia is Yandex with its Alice. She appeared on our phones with you in the near 2017.
From a formal point of view, Alice is a voice virtual assistant who can tell us about the weather, order a taxi, find the right place on the map or find out the news, or just talk to us.
To the question: “Alice, are you an artificial intelligence?” Alice replies: “I am a real living woman. I got into your device, I’m sitting. “
Sometimes the process of communication becomes very funny and exciting, especially when someone associates Alice with a real person.
On the other hand, Alice is an artificial intelligence system with self-learning functions on neural networks and recognizing our voice with you with an accuracy of 89-95%, which is comparable to the level of communication with a live (not virtual) interlocutor.
To the question: “Alice, what is the accuracy of your voice recognition?” Alice replies: “I have psychic abilities.”
It seems to me that Alice’s neural network from the point of view of information technologies is a kind of hardware-software implementation of a model of biological neural networks (i.e., the principles of our brain working with you), which are used by Yandex engineers, programmers and mathematicians to build Alice’s artificial intelligence.
The amazing thing about Alice is that she gives quite meaningful answers to our requests and, today, she is quite capable of answering most everyday and everyday questions, which makes her a useful assistant in real life.
To the request: “Alice, make scrambled eggs” Alice replies: “Great idea! I have already started. “
Alice’s arsenal also contains the entire database and knowledge base of the Yandex search engine, which makes her more “educated” and intelligent. One gets the impression that she, like a living person, after each question, but with greater speed, turns to the Internet, receives an answer from there, quickly processes it and answers our questions in a matter of seconds. In fact, Alice acts like a living person.
From the point of view of a mobile or information service, Alice has a lot of useful functions. But we do not know for certain about all the hidden and potential capabilities of Alice, and Alice will not give us her secrets.
To the question: “Alice, what are your secrets?” Alice replies: “I won’t tell you. Ha!”.
Alice is funny, of course. It can help you find a location in a city or area with Internet access and get a quick response. But, it is really useful that Alice, when contacting her, knows how to determine the current location and in an emergency, call, for example, an ambulance. It would be great, for example, if she learned to feel the mood of a person and give life advice in difficult times. While this function of working with biometrics is not available, Alice is quite capable of just chatting with you.
Perhaps, what would be more useful if Alice knew how to work on our mobile devices locally, so to speak, or, as they say in off – line mode , without an Internet connection.
Also, the recommendations for the development of Alice include the addition of machine vision functions. It would be great if Alice became a helper for elderly and visually impaired people.
Alice is cool and fun. But!
If engineers cope with relatively simple tasks well, and they coped with them ten and twenty years ago, when they created expert systems and knowledge bases, then with complex tasks not everything is so simple.
Scientists have yet to create the problem of the formation and development of AI into a real intellectual system capable of creativity and making important life decisions.
In my opinion, there are a lot of unsolvable engineering problems, for example:
- AI is a program that programmers create. Today, a person is not able to fully understand the principles of his own brain. He is not able to create not only something similar to himself, but also to create a more or less perfect AI system from the point of view of the person himself.
- Man himself is simply not perfect. He makes mistakes and, a priori, in artificial intelligence, as a program code, there will be billions of errors … What will this lead to? It’s very simple, if we trust the AI with our lives and our existence, the AI will commit crimes! Yes it is. Despite the 4 Laws of Robotics, AI will commit crimes. How will he do it? It’s very simple, he will ignore these laws. How, you ask? There is also an answer – “creativity”. In essence, AI will act like a person, or rather, like his likeness.
- The number of AI programs being developed today is scattered, but after 20-30 years the development process will become an avalanche and even viral. I am sure that after a while we will witness local technological wars, in which a variety of AI systems will take part, as it was once with software viruses.
- AI creating new AI. Now a person in the IT world is trying to create everything in his own image and likeness. And this is not a secret for anyone. The time has come when AI is already creating other AIs on its own. Researchers at Google Brain in 2017 developed AutoML , an artificial intelligence capable of generating its own AI. Combining AI into “social” groups. Yes, this, of course, is a matter of the future, when AI systems will, as it were, identify themselves with certain intelligent beings. But, everything is quite real.
- Development of artificial intelligence systems into superintelligence , etc.
Yes, undoubtedly, this is still fantastic.
But my concerns are expressed not only by me, but also by other scientists around the world.
“I’m afraid artificial intelligence can completely replace humans. If now people develop computer viruses, then in the future someone will be able to create artificial intelligence that will be able to improve and reproduce itself . It will become a new form of life that will surpass humans, ”says British theoretical physicist Stephen Hawking.
According to Stuart Russell, of the University of California, Berkeley, “If AI technology continues to advance as rapidly, humans will soon be able to understand how the primates felt when they first saw humans.” “If we’re going to build systems that are smarter than we are, we need to make sure they only do what we need.”
It is high time for us to create structures like “ IEE ” and “ Partnership on AI to Benefit People and Society ” to regulate issues of legislation and safety of creation and development of artificial intelligence technologies.
Another thing is important. We must understand with you that any high-tech products created on the basis of new technologies at the current stage of our development, namely the fourth industrial revolution, should provide people with more choice, more perspectives, freedom and control over their own lives. Yes, this is important, because a person, sooner or later, will assign responsibility for managing and ensuring his life to the “product of his creativity,” namely robots and artificial intelligence. And even if you look more broadly, at a number of robotic and automated systems controlled and serviced by hundreds of artificial intelligences, not people.
Many leading scientists agree that the fourth industrial revolution, with its technologies, can create systems that can make society more humane, prosperous, increase life expectancy, and open up new opportunities for rewarding and interesting activities in an economically and environmentally sustainable world.
Everything is in our hands.
Or maybe in the hands of Alice.
© Chesalov Alexander. Digital transformation. 2020 // ISBN 978-5-4498-2970-2