Business

As workers fear for AI job cuts, Open AI co-founder says AI agents will take a decade before they even work: ‘They don’t have enough intelligence’

For employees who worry that AI agents might take their jobs, one of the industry’s leading founders says the technology is still far from perfect. Andrei Karpathy, co-founder of OpenAI, said this is not “the year of agents” — and although he uses AI agent tools like Claude and Codex, they still lag far behind the work of humans.

“They lack cognitive awareness, and it’s not working,” Karpathi said in an episode of the Dwarkesh Podcast. “It will take about a decade to solve all these problems.”

He added: “They are not working. They do not have enough intelligence, they do not have enough multimedia, they cannot use the computer and all these things.” “They don’t have continuous learning. You can’t just tell them something and they’ll remember it.”

There is no isolated definition of AI agents, but they are typically used as virtual assistants that perform workplace tasks autonomously with logic. Today, agents are used for tasks such as customer service and IT support requests.

While many workers are becoming increasingly concerned about the security of their jobs amid AI disruption, Karpathy says the tools haven’t been perfected yet without someone to guide them.

“You should think of it as an employee or intern that you would hire to work for you,” he said.

The tech expert further elaborated on his argument for X, saying workers should learn from AI, not be marginalized by it.

“I want him to make fewer assumptions and ask/collaborate with me when I’m not sure about something. I want to learn along the way and become better as a programmer, not just have mountains of code that I’m told works.”

AI agents are still on the path to perfection, with administrative tasks at the forefront of the revolution

Today, AI agents are used for customer service, IT, and administrative tasks, but many tech companies are already scaling back their automation plans.

In fact, 50% of organizations that expected to significantly reduce their customer service workforce by 2027 are now abandoning those plans, according to Gartner, Inc. Also, 95% of companies that implemented AI pilot programs failed.

However, this has not stopped AI companies from trying to deal with these setbacks. For example, McKinsey built an AI agent using Microsoft’s Copilot Studio that can monitor the email address of project proposals from potential clients. While a human must verify what an agent produces, it has reduced the time needed to review projects from 20 days to two days.

On the other hand, in India, a company called LimeChat still insists on cutting customer service jobs, saying it will use generative AI agents to enable customers to reduce the number of workers needed to handle 10,000 monthly inquiries by 80%.

While the long-term impact of AI agents on the workforce is still unfolding, experts predict that they could significantly disrupt administrative tasks.

Don’t miss more hot News like this! Click here to discover the latest in Business news!

2025-10-20 15:30:00

Related Articles

Back to top button