Are AI agents ready for deployment?

Although automation holds great potential to transform the way work will be done in the future, the short-term reality for businesses is a different story.

The first generation of generative AI assistants and co-pilots have now become what Gartner calls the “in the depths of disillusionment“, with many projects remaining in the pilot stage due to a combination of factors: change management, lack of clarity on ROI and various security considerations, for example. And then there is the propensity of language models to “hallucinate” the answers. Many of the same challenges will be encountered when deploying AI agents.

Companies are understandably cautious about letting LLM-based agents act autonomously and access enterprise systems, for example, even if they are subject to limitations in terms of the actions they can take. are programmed to perform.

For now, most businesses will want some form of human oversight. “There is no circumstance, at least right now, where you would deploy this without a ‘human knowing,’” Kropp said. This means that human workers have visibility into the agent’s actions and are consulted before taking riskier actions. That said, Kropp is confident that the problem of AI hallucinations will become less important as the technology matures, with agents’ reasoning skills also rapidly improving.

While there will likely be many challenges along the way, Marsh expects the combination of AI and automation will ultimately have a “profound” impact on the way work gets done – even more so than other recent workplace changes brought about by new technologies.

“I think the productivity gains are there. I think they are real… If I think about all these changes that have happened over the last five years, this one will be by far the most significant,” he said.