Agentic AI – why agents must understand context to help enterprises align with goals
One of the challenges of the generative and agentic AI age is easy to state but harder to fix. Many of the transactional benefits of automating a task – writing a report, say, or summarising research – only accrue if you don’t investigate the output. But if you do spend hours checking every sentence, citation, statistic, and calculation you lose much of the time saved, while AI adoption increases the pressure to hit ever shorter deadlines, as I explained in my recent Monday Morning Moan.
However, at least one AI vendor believes that the problem lies in poorly trained systems, rather than flaws in AI models themselves. Such systems are the ones that create more work than they remove, claims contextual agent specialist Orient, leaving employees trapped in a cycle of “verifying, correcting and contextualising” slop.
By extension, this is why organizations struggle to see any ROI, says Orient,...
Copyright of this story solely belongs to diginomica.com. To see the full text click HERE
