At the mobile world conference, the commercial exhibition for the communications industry is held every March in Barcelona, The agents of artificial intelligence were everywhere. Or rather, the signs promoting artificial intelligence were everywhere. in Google Cloud Stand, Telecom Execs can see illustrations that show how easy to build dedicated AI agents. in Microsoft Stand, it was the same story. Qualcomm’s It highlighted “Ai Agenic Ai”.

It was a similar scene in January in World Economic Forum in DavosSwitzerland – where the offers of technology companies “Agency Ai” that practically obscured the vision of the mountains that the surrounding snow holds. In fact, all technology companies turn almost a place to market their goods these days into the Ai-Agent Festival.

Simply, there is no more sincere topic in technology and business now. Artificial intelligence agents are supposed to be finally delivered from artificial intelligence Productivity gains that have long emerged. The first days are still cautious, and this is partly due to Artificial intelligence agents carry more risks than other artificial intelligence products. But if the publication is quickly intensified, many analysts expect, Artificial intelligence can radically turn how people work– It may lead to cutting millions of jobs.

But what are the factors of artificial intelligence, and is it worth the noise around them? For beginners, there is no agreed definition for the artificial intelligence agent-which is appropriate if you want to claim one sale. But in general, it is a system that works with the same Amnesty International that can complete tasks using other software tools. These systems contain the Generative-AA model in essence, but they can do more than the artificial intelligence model that can be in isolation. The GEN-AA model may be able to do so I suggest a line to spend a next holiday. The artificial intelligence agent can do this as well, but then In fact, make reservations and reservations for you. It is clear that this is a more difficult task that includes multiple steps and the use of the Internet – and a good job, a great deal of thinking.

This type of automation is a permanent c-hite fever dream. Over the past decade, companies have adopted “automation of automatic operations” or RPA. This was a program that could Automation of repeated tasksLike cutting and paste between database programs. But traditional RPA systems are unable to deal with exceptions, and they can usually deal with only one narrow task. Amit Zveri, head of the Business Program Company, says that RPA has not provided what companies want. servicenowAnd it is the great bet on the agents. Agency AI says: “It can be flexible and adapt to how companies work.”

The American market for artificial intelligence agents is expected to reach $ 7.6 billion this year, according to Grand View research, and this number grows by 46 % annually. But these numbers are likely to reduce the total effect. Consultants in McKinsey Suppose Amnesty International can eventually open up to $ 4.4 trillion of annual companies ’productivity – with a lot of that next from AIC.

Currently, adoption is still arising. Roshir Buri, chief scientist says IBM research. Meanwhile, Craig Le Clere, an analyst at the Forster Research Company and author of the book ” Random automation actionsA book about the next world for artificial intelligence agents says that there are many “agent” offers, however A little independent factors are really independent Their approach can be adjusted to a task while flying in response to changing conditions.

Currently, agents are often used to serve customers and solve internal information technology support requests. However, it is not yet good enough, with automation of tasks just like writing programs and errors, says Buri.

Since agents can take measures, such as managing critical databases and participating in financial transactions, the risks are high if things get worse. Thus, the major companies are cautious about technology adoption.

However, agents sneak into many areas – from helping with new employees to help lawyers find the language and update it in contracts.

In a pilot project, MCKINSEY has built Amnesty International’s agent using the Microsoft Copilot Studio program that can monitor an e -mail address for project proposals from potential customers. When one reaches the incoming box, the agent automatically evaluates the job, estimates employment requirements, ends, and the budget. He even suggested that the advisors available to do this task. A person still has to check what the agent produces, of course, but technology has reduced the time required to determine a 20 -day project on average to only two days.

Double agent

Approval on artificial intelligence agents rises, but many fear they will kill jobs

32 %

The share of workers who believe that artificial intelligence will lead to fewer jobs

7.6 billion dollars

We are expected to spend on artificial intelligence agents this year

Sources: Pew Research Center; Grand View Research

What is the effect Artificial intelligence agents on workers? “I think it will be really annoying,” says Lu Claire. It knows an insurance company in the Netherlands, which has 15 contractors in Bulgaria, helping to process the calls related to demands. Contractors decided whether an email to a new claim or an existing message, and make sure to download any information, including attachments, to the appropriate databases. Artificial intelligence agents are now working – and the company launched all Bulgarian contractors.

But such cases may be the exception, at least at the present time. A recent study conducted by two senior banks of international settlement consultants found artificial intelligence agents. “They lack self -awareness to know when they may make mistakes and change the path in light of the evidence,” although they lack self -awareness to know when they may make mistakes and change the path in light of the evidence. “People also make mistakes, but they are better in discovering and correcting their mistakes.

Of course, artificial intelligence companies are racing to overcome these restrictions. The new wave of thinking models such as Openai, Human, Google’s DeepmindThe start of the Chinese Deepseek is outperforming step -by -step plans and have some ability to think about these plans and make corrections. Adam Evans, Executive Vice President who heads sales efforts of artificial intelligence, says the latest artificial intelligence agents, which were released in March, began to integrate some of the capabilities of the causes. The company is also working on ways to help companies organize many agents to deal with the most complex tasks.

So, although most of our jobs may be safe for the next year or two years, we may all find ourselves and soon we scream that the old Hollywood Shippolith – “called my countryman!” To re -negotiate our employment contracts.

This article appears in April/May 2025 version to luck With the title “Artificial Intelligence Agents here. How much workers’ fear?”

This story was originally shown on Fortune.com

https://fortune.com/img-assets/wp-content/uploads/2025/03/CB_AIAgents_Web.jpg?resize=1200,600
2025-04-01 10:00:00