Development of neural networks can threaten people
Recent years, the development of neural networks is very fast. This cannot but alarm, since any technology is both a benefit and harm to humanity. What awaits us further with the development of AI?
One of the experts, Yevgeny Volkov, is convinced that even now a person cannot consume all the information that is poured on him from all sides. Already rarely meet such a person who would read the newspaper from the first line to the last – something is sure to distract him. Brains do not cope well with this, and children already have an innate media filter.
Artificial intelligence acts as a cognitive assistant for a person, relieving small informational messages and notifications, helping to make the necessary decisions. In the future, according to Volkov, these smart assistants will be combined into a separate network and interact with each other. So, when you say aloud “you could meet with Anya for a cup of coffee”, your virtual assistant will contact the assistant Ani, they will pick up a coffee shop where you like, based on your history of visits, book a table and send you a time notification and place. Maybe even they make an advance payment. Volkov believes that it is real now, but everywhere such services will appear somewhere in 3-5 years.
Development of neural networks and matter of privacy
If while you are sleeping, your smartphone is next to you, then Google knows how well you sleep, how long, with whom (if your phones are connected to the same wi-fi network). People continue to defend the confidentiality of their data, not wanting to admit that they have already lost this war. Volkov is convinced that in the future virtual assistants can even be our personal HR-managers, because they know much more about us than our own parents, and, accordingly, can better recruit staff to solve any problems.
Despite the fact that now the process of learning neural networks takes a long time, at the end of this process, decision-making takes much less time from artificial intelligence than from a human being. Neural networks surround us, help us perform routine tasks hourly, but the problem is that even at this stage we cannot understand exactly how they make decisions, on the basis of what. A human can process 5-10 parameters, a neural network – 2000. Even now, developers themselves from Yandex cannot precisely explain why a particular website goes to the top of the list.
This is done by artificial intelligence based on thousands of factors, without even notifying the people responsible for it.
The same is observed in many other spheres of activity, therefore Volkov recommends all of us to think about what we will do in 5 years. Automation will hook every sphere of our life, and many professions that today seem to be absolutely “human” will quickly go into the past.