Meet your new AI workplace companion

Will 2025 be the year AI assists humans at work? If we are to believe Jensen Huang, NVIDIA’s CEO, the answer is yes. During his recent two-hour keynote at GTC, he laid out a vision of a world in which agents are everywhere.

“This is the way enterprise will work in the future,” he said. “We’ll have AI agents [that] are part of our digital workforce.” By the end of this decade, he added, the world could possibly see a shortage of human laborers. “We're probably going to have to pay robots USD 50,000 a year to come to work. And so this is going to be a very, very large industry.”

As companies like San Jose’s Figure, Amazon and Tesla work to make home robots a reality, startups building AI agents for the enterprise are also popping up. Roam is building virtual HQs for workforces—human and agents alike—promising to “link people, offices and AI agents to work side by side to create a bustling virtual HQ.”

Artisan, another startup, builds AI virtual employees for outbound processes. In San Francisco, the company's ads have started popping up on bus stops and buildings, and have also drawn attention—and drove controversy—in recent months. “Stop hiring humans,” the tagline says. Artisan recently raised USD 25 million in Series A funding.

Humatron AI is another new AI workforce-centric company. “Our idea is pretty simple,” says Humatron Founder Nikita Ivanov, in an interview with IBM Think. “In our platform, you can build workers—you can build for yourself, for your company, you can build for public consumption. You can basically build a worker … and anybody else can hire them.”

The idea is to make the hiring process of a specialized agent similar to what enterprises already know, from interview to onboarding to pay. And this digital employee will be more than just a bot. “You shouldn't see much of a difference between that AI work and a human colleague in day-to-day work,” says Ivanov. “They're in Slack, they're in email, they're on Zoom—they're everywhere.”

For businesses, AI agents are already opening up new opportunities, says Karen Butner, a Global Research Leader for the IBM Institute for Business Value (IBV), including by accelerating business operations automation. In a survey conducted by IBV for a new report, 60% of executives said that their employees would be interacting with AI assistants by the end of the year. Agentic decision-making will also be on the rise, according to executives surveyed, and 70% said they believe that AI agents will enable business professionals to drill deeper into analytics to support real-time analysis and optimization by the end of 2026.

“Think about 24/7 operations in a business, no matter where it is, across the globe, all the different time zones, translated into all the different languages, happening like this—and people aren't having to make the adjustments and say, ‘Oh, this product's going to Brazil. We have to provide the proper language and labeling adjustments,’” Butner says in an interview with IBM Think. “The agents are able to make those calls—and a whole lot more.”

In a recent blog post, OpenAI CEO Sam Altman predicted the rise of agent coworkers as one of the ways AGI will manifest. “We are now starting to roll out AI agents, which will eventually feel like virtual co-workers,” he wrote.

IBM Distinguished Engineer Michael (Max) Maximilien can also imagine a future where agents are part of a dev team. “Instead of having a team where, let's say, one person is taking requirements, one person is doing the UI, one person is doing some of the engineering, one person is doing some of the testing and then one person is doing the DevOps to deploy—I could see that the DevOps and the testing are being done mostly by AI,” he says in an interview.

Are AI agents going to replace human workers? Ivanov, from Humatron AI, predicts that humans will still make up most of the workforce, though companies could use AI agents for entry-level positions. “It's an interesting and not-so-simple dilemma,” he says, referring to the concern of AI taking people’s jobs. “What do you do? Well, the only way to solve that is to really change the curriculum—how we prep and study for these jobs.”

In a recent blog post, Ann Funai, CIO and VP for Business Platform Transformation at IBM, asserts that AI agents are meant to augment humans, not replace them. “[It] is not about humans versus machines,” she wrote. “It's about humans and machines working together to achieve better outcomes.” However, the technology still needs to be used correctly to truly capitalize on its potential, Funai added.

When it comes to the future of work, Maximilien says it’s still hard to predict what’s next, because the capabilities evolve so quickly. “If you look at GPT-4 or DeepSeek and you compare it with what we were looking at last year this time—it’s like leaps and bounds. What’s going to happen next year?”

“I think AI is one of those unique technologies that has the potential to impact every part [of work],” says Maximilien. “For the longest time, we always kind of prided ourselves on being better than machines. But now, the genius of a lot of these AI models, when you train them correctly, is that they can come up with answers to things that you would never expect because they understand language.”