"GenAI has an error rate of up to 25 percent depending on the use case. That might be OK in some areas, but certainly not in others. The problem is that 84 percent of CIOs and IT leaders "don't have a formal process to track AI accuracy", said Rob O'Donohue, VP analyst at Gartner. The idea of mitigating this by having a human in the loop, as some vendors still cling to, is unsustainable, as AI output is more than humans can keep up with."
"A Gartner survey of 700 CIOs indicates that, by the end of the decade, all business IT work will involve AI, while bots will do 25 percent of that work by themselves. Good news: The analyst firm claims AI causes only one percent of job losses. Bad news: You'll have to learn some new jargon. To survive the transition, it might be helpful to at least sound like you are up on the latest terminology, even if you can't bring yourself to use it."
"The pitch from many a vendor is that agentic AI can handle the more mundane aspects of a team's work, allowing its members to focus on higher-level things. The problem is that those very mundane bits of work are where junior members of staff once cut their teeth. Okay, so you don't need to hire so many staff out of college or second-jobbers."
A Gartner survey of 700 CIOs projects that all business IT work will involve AI by the end of the decade, with bots performing 25 percent of that work autonomously. The firm reports AI accounts for only one percent of job losses, while new terminology and practices will become necessary. GenAI can have error rates up to 25 percent, and 84 percent of CIOs lack formal processes to track accuracy. Relying on humans-in-the-loop is unsustainable given output volume. Organizations need an AI accuracy survival kit—formal metrics, two-factor error checks, and model-to-model validation—and must address junior hiring and succession risks from agentic AI adoption.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]