Business loads neural networks – Kommersant FM
[ad_1]
Artificial intelligence has proven to be an effective tool for US sales managers, saving employees more than two hours every day. This conclusion was reached by the American developer of marketing products HubSpot. The company surveyed almost 650 sales professionals and more than 300 business people. It turned out that most of them use AI to perform routine tasks – scheduling meetings, taking notes and entering data. In Russia, such programs are most often used in marketing, customer service, development and IT. In December 2023, experts from Yandex and Yakov and Partners came to this conclusion.
Kommersant FM found out from business representatives whether they use artificial intelligence and what kind of work they are willing to entrust to it. Luka Safonov, CEO of the information security company Cyberpolygon, said that he has been using neural networks for several years, mainly to solve simple problems:
“We got acquainted with neural networks quite a long time ago, when it was not yet on the wave of hype. They are used in some simple things, for example, writing autotests or some forms. That is, you can upload a piece of code into a neural network, if the program allows it, and ask the AI to write autotests. And all the employees were kicked out into the cold. I’m kidding, of course. They, among other things, write prompts for neural networks and so on.
Machines have not yet reached the level where they can do absolutely the entire task without human control, because there are a lot of “buts”. Something may go wrong, the algorithm may be disrupted, some external intervention may occur where human action and experience is necessary.”
In retail and industry, artificial intelligence can now be useful both for communicating with customers and for solving more ambitious tasks, says the owner of the furniture manufacturing company Mr. Doors Maxim Valetsky. He plans a large-scale implementation of neural networks in the near future:
“In fact, in 95% of cases, artificial intelligence is simply understood as a neural network that finds the optimal path to solve a problem and learns something. We are going to use them in design. This is a long process, there is no ready-made solution yet, but we are working on it. We will gradually use what is closer to intelligence when texts are created in marketing; we can try such a tool in communication with customers in CRM systems.
But today a huge problem has arisen in this area: artificial intelligence is now creating meanings that do not exist.
The most advanced systems, such as ChatGPT, have already absorbed all verified sources. To educate him further, he is fed posts from social networks. Thus, a gigantic amount of information has appeared that the AI itself generates and which is simply not reliable.”
According to Vedomosti, in Russia artificial intelligence is also actively used in transport companies. For example, oil and gas field operator XP Group has been using machine learning since 2019 to improve demand forecasting, logistics and assortment analysis.
True, market participants are not ready to entrust some processes to neural networks, says Kirill Latinsky, executive director of Transasia Logistics. For now, he is only thinking about using artificial intelligence in the company: “We are looking in this direction. You cannot completely trust AI from a recruitment perspective. After all, this is the language of numbers; he definitely lacks any feelings and emotions.
You should also be wary of such a tool from the point of view of resolving any conflict situations both within the team and in relationships with clients.
AI cannot be used in the process of payments through a client bank to our contractors. Also, this tool cannot be used from the point of view of settlements with government agencies regarding taxes and fees. At the same time, there are a number of important business processes where the help of artificial intelligence would be indispensable. In particular, we are talking about tracking the dynamics of changes in operating profits, taking into account the use of third-party financing, due to the fact that it often has floating interest rates that are tied to a key indicator established by the Central Bank.
If we can see in real time, from an open period, how production profit changes taking into account various influences on the transportation itself, then that would be great. It would be possible to refuse some transportation in advance, reduce costs in some places, and, on the contrary, increase tariffs in others.”
The Ministry of Internal Affairs also plans to turn to artificial intelligence. As Vedomosti wrote, in 2025 the Ministry of Internal Affairs will develop two systems based on a neural network – “Clone” and “Conjuncture”. The first is expected to make it possible to identify fake videos, and the second will be able to predict negative events and emergencies, and then model response scenarios to them.
Everything is clear with us – Telegram channel “Kommersant FM”.
[ad_2]
Source link
