Can machines think? This question takes both developers and philosophers. British mathematician and logician Alan Turing suggested that the mind is just an algorithm. And this algorithm can perform not only our brain, but also the computer. On this basis, Turing concluded that any machine that would be wise to conduct a dialogue on natural human language, will actually think. So the idea of artificial intelligence.
Technology development and application
Historians believe that the end of XX and beginning of XXI century covered three waves of technology development of AI. The first wave — a naive attempt to create “General intelligence” through formal rules of logical thinking. Second, the development of “expert systems”. We are now at the crest of the third and most powerful wave, and it is connected with computer training. Programmers have learned to create systems, which are themselves improving algorithms and get new data. This invention created a new industry. Researchers from Morgan Stanley believe that it is growing annually by 16% per year, and by 2050 will reach $1 trillion. The introduction of technology computer training to automate and greatly improve the results of human labor. In addition, experts from McKinsey, it will add trillions of dollars to global GDP.
This will make a profit in the first place, tech giants such as Google, Facebook, Uber, Apple. With the help of computer training they will be able to solve the difficult task of creating computer vision modules, motion control, speech understanding, organizing, and providing access to information. But for the smaller parties of the third wave has excellent prospects. They are the application of artificial intelligence to specific local problems. I will cite just a few areas where I think the prospects for the use of technology is most apparent. In these areas it is now possible to create successful startups, not having huge start-up capital.
The first of them is the safety of the house. Many of us installed home cameras. But how often do we actually look at the archives video? I personally even forgot the password to them. Viewing these videos is a long and boring process. I wish to entrust it to someone else! Artificial intelligence would be an excellent choice for this role. To find a use for security trying a company. Netatmo Welcome and Nest Cam IQ has taught the camera to recognize faces. And the developers of the Lighthouse and Cherry Labs taught algorithms to analyze home video and identify events in the house, violating the usual routine. For example, the nanny left the child unattended, plumber instead of food came to you in the bedroom, in the house of an unknown person. All of these situations the artificial intelligence can identify and send you a text with a link to the video. And then you yourself decide what to do with it.
In addition to these custom events, home artificial intelligence can see charm in casual events. The developers told me that during testing, the program saw an unusual how an elderly woman sits on the couch. They decided that it was a glitch, and the program just did not understand that this is the most ordinary event. But it turned out that the machine was right: the older woman didn’t sit down, and fell on the couch. The sofa was just too low, and the program noted the unusual speed of this movement. Had to tighten up the legs, to home intelligence calmed down.
Another area of the most promising applications of artificial intelligence medical diagnosis. Russian and Belarusian developers made the program capable of using neural networks to accurately predict the time of the female cycle. It turns out that this is a serious problem: more than 30% of women in the world have a regular cycle, it depends on many things in lifestyle and planning for posterity. Users of the app Flo for a period must record the parameters and events of the organism, and, after some time, the program is better than any doctor will know what will happen and when. These data are useful to prevent pregnancy, or conception, and to detect diseases that require the attention of these doctors. Of course, there are ethical issues. For example, if we have the right to collect data about women’s health? Flo collects information even about sexual contact, and it is very private question. But the ethical issues apply to absolutely all cases, the development of AI.
Women’s calendars are numerous, but those that use neural networks to predict, as far as I know, no more. In addition, this application went over the top in its category. The success of this startup is difficult to overstate: it has received over one million five star reviews from users, has become the second most download app in category Health and Fitness in the Appstore and raised a round of $200 million from a leading European venture Fund Mangrove Captial Partners. The company has reached the breakeven point and may soon become a unicorn.
In addition to artificial intelligence, and machine learning will radically change the management of the organization and take on what is traditionally considered difficult psychological work of the HR function. This task put before itself by the company YVA.AI. They recently successfully closed the investment round and was in the shortlist prize “Made in Russia”.
Service Yva allows you to take another look at the staff of the company. The program connects to all communication channels (instant messengers, mail, phone) and analyse how employees communicate with each other, with partners and customers. After that, she is able to draw General conclusions about the quality of work of individual employees and departments as a whole. She “knows” who is influencers who initiates change and who passively reacts to them, who really keeps track of tasks where there are conflicts and potential problems. The service, for example, allows you to discover the best seller in the company and behaviour patterns: this knowledge can be taught and other sellers. With high probability it can also predict the future discharge of an employee, even if he has not yet taken such a decision. Perhaps in the near future, such algorithms will be a crucial assistant Manager in the company and be better than him to make decisions who to fire and who to promote up the career ladder.
Philosophy of technology
We see how artificial intelligence technology can become the core of the business and how they can make more comfortable our lives. But it remains an important philosophical question: can artificial intelligence really thinking? Will we ever be considered advanced versions of the algorithms CherryLabs, FLO or YVA implemented, for example, in robots, the real thinking beings? whether we consider them as our security guards, doctors and HR consultants?
Among our contemporaries there is one philosopher who denies this possibility in principle. Is the American John Searle. In his opinion, no matter how intelligent did not seem the behavior of computers, they will always be just advanced calculators. The argument he came up with a hypothetical situation, which became known as the “Chinese room.” Imagine a room in which sits the English speaking people. Through the window pass him questions in Chinese, and he needs to answer them. And although he does not know Chinese, he has an infinite number of guides and instructions that specify how to form the answers to this incomprehensible language for him. Guided by these rules, a person records the answers and sends them through the window, not understanding what he writes, since the text is formed according to the instructions without translation. But the Chinese outside the room, the impression that he understands their language.
Searle believes that a similar situation describes the intelligent behavior of computers. Perhaps, the computers transmit the answers to other observers seem to be meaningful. That is, apparently they can behave reasonably, to give us advice or even to maintain a dialogue, as suggested by Alan Turing. But at its core, the computers will never become intelligent. At a basic level, they operate on only ones and zeros recorded on the hard disk, and the instructions for changing these numbers, which are defined by programmers. Neither these data nor instructions not enough to create understanding. The consciousness, or connection to the world, which have only people. It is a real, genuine thinking.
When Siri will be able to converse like a human?
Read more •••