The past two years have been the start of the era of artificial intelligence, and the prevalence of AI in workplaces may be increasing faster than people realize.
AI is advancing at a breakneck pace, especially in the field of generative AI. What started out as a tool that could help the everyman in their tasks and improve efficiency is continuing to refine itself, and now, the leaders in the AI field are already touting what can come next.
If the predictions of top AI industry figures like OpenAI's Sam Altman and Meta's Mark Zuckerberg are to be believed, 2025 is set to be the start of the era of the AI super-agent. These AIs are, in theory, smarter and have far more utility than their predecessors, and are capable of carrying out more complex tasks by themselves.
To put it simply, AI super-agents could potentially be able to do jobs typically done by humans, meaning AI may be coming for your job.
But what is the truth behind this claim? What is pushing the public discourse around AI super-agents? And will something like ChatGPT really be coming for your job in the near future?
Here is everything you need to know about AI super-agents and what they can, in theory, do.
What is an AI super-agent?
To understand what a super-agent is, it's important to first understand what an AI agent is.
Putting it simply, an AI agent is an AI software that can work towards a specific end by itself. This means it can perform tasks without human assistance and can make decisions on its own.
This itself isn't anything new. Other companies have already debuted AI agents that can interact with computers, and OpenAI has already been reported to be working on an AI agent that can interact with a web browser by itself. That AI agent, reported to be called "Operator," according to reports published back in November 2024, could debut before the end of January 2025, or at least a preview of it, but what exactly it can do and what its limitations are have yet to be revealed.
A super-agent, in theory, is an AI that does even more than that. According to a report from Axios, rather than responding to a simple command or undertaking simple tasks, super-agents would be able to carry out far more complex matters in the pursuit of their goal. This doesn't just mean making decisions, it could mean creating complex and detailed plans and analyses to plan what is essentially an entire vacation, including making the relevant bookings and reservations or even creating entire products from scratch.
Will AI super-agents be coming for your jobs?
This is the biggest question with new advancements in AI, and for good reason. In the past few years, many companies around the world have already seen widespread adoption of AI, which has come at the cost of human staffers.
According to a June 2024 survey by Duke University and the Federal Reserve Banks of Richmond and Atlanta, nearly 60% of companies have already started implementing ways to automate tasks that human employees used to do. And according to that same survey, that is set to continue into 2025.
The reason for this is twofold. AI being automated can be used to boost productivity and efficiency, and AI is also cheaper than humans.
But if an AI super-agent is as good as many are predicting, AI could see even more widespread use in a wide range of industries.
For example, making a product, such as software, from scratch is a legitimate possibility for an AI super-agent. In fact, Mark Zuckerberg said as much in a recent appearance on The Joe Rogan Experience podcast. There, he predicted that Meta and other companies could soon release an AI that "can effectively be a sort of midlevel engineer that you have at your company that can write code" and that this could be a reality in 2025.
OpenAI's Sam Altman hasn't been as vocal about this timetable as Zuckerberg, but he and his company have been touting the prowess of the next generation of AI, with OpenAI officials having compared them with a human with a PhD. In addition, Altman is also set to meet with US lawmakers and officials to demonstrate OpenAI's new tech, which will "show the economic power of AI," according to The New York Times.
Will AI super-agents really be coming for your jobs in 2025?
Despite the bold claims in many reports about the imminent arrival of AI super-agents, it is unclear how much of what is being touted about these yet-to-be-released programs is genuine rather than simple hype.
Taking to X/Twitter, Altman said that Twitter hype was "out of control" and stressed that the company is not going to deploy artificial general intelligence (AGI), a term used for AI with cognitive skills on par with or greater than that of a human, within the month, and stressed that people's expectations should be lowered.
twitter hype is out of control again. we are not gonna deploy AGI next month, nor have we built it.we have some very cool stuff for you but pls chill and cut your expectations 100x!
— Sam Altman (@sama) January 20, 2025
The possibility of AI taking over many jobs, however, is still something many experts say is likely, but not in the immediate future.
Speaking to CNN in June 2024, LinkedIn co-founder Reid Hoffman said that AI disruption in the labor market is coming, but not right away.
"I believe in three to five years, we’ll all have kind of an agent co-pilot that's helping us with anything from how we cook dinner… to doing your job and writing and so forth," he told CNN. "It's job transformation. Human jobs will be replaced - but will be replaced by other humans using AI."