"Everyone will be a manager" - what the world of work will look like with AI agents
"Everyone will be a manager" - what the world of work will look like with AI agents
AI agents are reshaping roles, empowering employees, and redefining leadership in the workplace.
The ground in the world of work is shaking. Last week, Mark Zuckerberg, founder and CEO of Meta, said that he will replace junior programmers with artificial intelligence as early as this year. He made the remarks on an episode of The Joe Rogan Experience podcast. A few weeks earlier, on another podcast, 20VC with Harry Stebbings, Salesforce CEO Marc Benioff announced that the company would not hire new software engineers in 2025 due to the improvement in productivity thanks to the use of AI. He said that the company is now focusing on its flagship product in the field of artificial intelligence—Agentforce, a platform for creating artificial intelligence agents. Just to put it in perspective, a mid-level engineer at Meta in the U.S. earns an annual salary in the six figures, so the savings for the company could be significant, and the implications for the labor market are clear.
In October, Microsoft launched AI agents for the workplace, allowing employees and organizations to create customized agents that can perform tasks without writing code, using natural language. At the same time, Salesforce launched Agentforce, an AI agent designed for organizations, especially for sales, marketing, and commerce departments. Google also launched “Deep Research” a few weeks ago, which is essentially an AI agent that acts as a research assistant that investigates complex topics on the employee’s behalf and provides a detailed, easy-to-understand report. It creates a research plan that consists of several steps that can be edited or approved.
Despite the announcements and tectonic shifts, artificial intelligence experts keep saying that it will not replace workers but will complement them and that more jobs will be created thanks to technological advances. The AI revolution is here, but what exactly does it mean for each and every one of us in the labor market? What will it actually look like when AI agents work for us, and are they really here just to complement humans?
First of all, what are AI agents?
AI agents are advanced artificial intelligence systems designed to perform complex and autonomous tasks. Unlike basic models like ChatGPT, or Google’s Gemini or Cloud, which provide answers to questions and respond to prompts reactively, AI agents operate proactively. They understand context, learn the work environment, and perform automated actions tailored to the specific needs of users.
"AI agents are divided into two main types—personal agents and business agents. Personal agents are tailored to the user's individual needs, such as an agent who can organize a trip for you—from flights to hotels. Business agents operate in organizations and streamline processes, such as answering employee questions or automatically performing HR tasks," says Dr. Tomer Simon, Chief Scientist at Microsoft Research and Development Israel.
The great news about AI agents is their ability to act proactively. They don't just wait for commands (prompts) but perform actions automatically. For example, instead of waiting for the employee to ask the AI to schedule a meeting, the AI agent can read the relevant email that requires a meeting, analyze the calendar, suggest alternative dates, and even manage coordination processes with several parties required to be in the meeting.
AI agents combine artificial intelligence technologies with organizational information infrastructures so that they can 'talk' to customers or employees and thus conduct service or sales calls or internal organizational coordination. "Agents today know how to access organizational information, draw conclusions, and perform actions," says Hadass Admon, Director of Product Management at Salesforce Israel.
For example, a service representative who takes a long time to read historical information about a customer and summarize the conversation can use an AI agent to perform these tasks in seconds. "The agent also helps human representatives by making recommendations on the work process. It can retrieve knowledge that the representative did not even remember being trained on, which improves their performance."
But it's pretty clear that the next step is for the agent to talk to the customer alone, and then what is the representative needed for?
"What will happen is that the work of the service representatives will change. We implemented Agentforce, and since then, from 10,000 inquiries that reached a human representative out of 36,000 inquiries per week—the number has dropped to only 5,000. So while there is supposedly less need for service representatives, we need to see what new roles have been created—perhaps we need to document knowledge or develop new automations. I read in studies that there is anxiety about the rise of artificial intelligence, but in my opinion, the change needs to be connected. Every team of service representatives always has more tasks than resources. Companies recruit endless service representatives, but what happens to the organization is that it has poor customer service. Therefore, organizations today are turning to artificial intelligence to streamline the service together with the people and build trust," says Admon.
And Salesforce itself is an example of how AI creates jobs. "Salesforce released Agentforce but is hiring 2,000 people to sell the system. It does free up tasks but still needs humans to do the more complex tasks," says Dr. Omri Allouche, Chief Scientist at Gong.
There's a lot of buzz around the idea that AI agents will replace people, but as they are educated about the tasks that employees perform, it's pretty clear that no one is hiring a salesperson to enter data into systems or to draft emails. "We hire salespeople to understand that who is the relevant person, and this is the pain that companies have that we believe we can solve. Everyone has to adopt these tools because if you don't adopt them, you'll be drowning in tasks that agents can do," he says.
Okay. So AI agents won’t replace sales reps, but given that AI can write code and now AI agents can be proactive in performing tasks, how do those agents help programmers and not stymie them?
"A digital workforce isn’t about replacing people—it’s about empowering them and allowing them to focus on the work that really matters. AI agents aren’t replacing programmers or other workers; they’re empowering them with tools they didn’t have before, allowing them to do their jobs more efficiently and focus on the more important tasks, thus increasing productivity, which currently allows the organization to continue with the existing workforce.
"Autonomous agents will change and impact almost every job in some way, and businesses will need to adapt and rebalance their workforces. Ultimately, these innovations will create more opportunities for us humans, and we need to take active action today to improve our skills to take advantage of these new opportunities," says Admon.
"Even today, if I take Copilot as an example, it helps me understand code and write code. A lot of the developer’s job is routine technical tasks like code analysis or writing basic code. If I can now have a developer focus more on critical tasks like system design, architecture planning, information security, and performance—and not focus on the technical part—that’s great, it’s really empowering," says Dr. Simon.
"Working with agents not only changes my work but also management because if my employees use AI agents, it means that everyone becomes a team leader, and then every employee is a manager. And then what does that say about me? As a manager, I can now deal less with bureaucracy and focus on more strategic processes, on making more substantive decisions and less on putting out fires. This way, organizations can be much more effective and focus on growth—from running the business to growing the business."
In addition to management, what will certainly change in the world of work is the way juniors are trained because many of the tasks they perform are well performed by AI agents. "Humans will have to do the more complex operations, and we will have to see if these basic skills should be taught or skipped. Will they start running before they learn to walk?" says Dr. Allouche. "What happens when using AI agents is that humans are put in the role of supervisor. AI agents do not replace human creativity but rather enhance it. They offer many ideas, most of which are irrelevant, but over time the quality of the ideas improves, and the real creativity remains with humans. Their ability to understand complex situations and solve problems with little data is still unique to them. The great news is that AI agents act like smart specialists, but to maximize their potential, we must become mentors who direct them, define rules, and guide them on how to act. In the future, AI will allow people to dream and create more, and human initiative will leap several levels forward."
