Translated material
"It all started with the fact that we were looking for nipples on the udder of cows"
“The neural network has learned to determine heart disease by the eyes of doctors,” we read the next news about the successes of artificial intelligence, almost without thinking about the processes behind all these achievements. At the same time, in order for the neural network to “learn” something, it first really needs to be taught this. And only a person can do this, or rather, thousands and even millions of people - only more than 500 thousand work on Amazon. First, you need to set some kind of pattern-for example, manually jump millions of images (a person in a picture or an animal?) Or, if there is already an array of data, mark it-structure disparate data by categories selected by researchers.
Not only scientists from universities and scientific centers are engaged in artificial intelligence - systems based on AI are gradually being introduced into business. Every major technological company has its own achievements - in the USA, for example, the Google and Facebook laboratories compete, in Russia - Yandex and Mail.ru. In addition, the introduction of AI may be required far from IT companies that do not have the opportunity to open individual units; Then they give these tasks to the side.
The Russian startup DBRAIN is a platform that combines customers from business, Data managers (Data Scientist-a person who offers the solution to the desired problem in the field of AI) and the final performers of neural networks. The creators of DBRanin say that, on the one hand, they reduce the threshold for entering this new area of technology for business, on the other hand, they give a source of earnings to low -skilled workers, at least partially removing nervousness against the backdrop of conversations that robots will soon deprive all jobs.
The DBRAN technical director Alexei Khakhunov , who participated in the creation of several other companies in the field of artificial intelligence, told how neural networks are changing business and labor market, as well as whether they can ever write a column as Oleg Kashin or other authors of Republic.
Unconditional neuromroduction
- When they talk about artificial intelligence and its influence on the labor market, they often represent a gloomy future: AI dismisses everyone from work, unemployment, chaos; In general, Skynet won. Is that so?
- Let's ask ourselves first to ask ourselves: does total employment need in principle? Perhaps we will come to the fact that jobs will become much smaller than people who need to work. Already several countries, including Sweden and Finland, are testing unconditional income.
- One of the favorite topics , yes.
-My personal offtopus is that the population will share 80% and 20% known by the Pareto law. 20% of people who are at high and who want to achieve something, and 80% are people who are now working because they need to work to eat something. They will simply tell them: “Guys, do what you want - look at KVN, walk or draw, why do you need work?” They will be happy to leave if they will live on. In almost all tasks related to machine learning, there is a need to mark data. This is such a simple monotonous work - those people we are talking about are perfect for it.
Our social goal is to simply give this monotonous work to people who have nowhere to work now, since they were born in the wrong place and not at that time. Moreover, we are talking about real, significant means. Now the cost of marking is about $ 3-7 per hour. On this you can safely earn $ 400-600 per month.
-Not being an IT specialist?
- Yes, we are talking about very simple tasks. If a person has any special knowledge, then the company will be able to pay him not six, but $ 20 per hour. People who, conditionally, work at the factory at the machine and are afraid that they will be fired is one class of problems. There is another problem when a doctor who studied at a university for six years, and then has a practice, is afraid that artificial intelligence will replace him. We can invite him to mark the data, for which it is precisely his special knowledge. We will give the opportunity to give part of the future profitability to people who will invest their ability in these specialists. Thus, doctors marking cancer will be able to become part of the future of AI.
Subscribe to read the whole
Make a subscription Redefine.media to read Republic
Subscribe [you can pay for a Russian or foreign card. The subscription is extended automatically. You can unsubscribe at any time.]
Where are the money subscribers going