Pretty Soon, GenAI May Have Conversations About You Behind Your Back

This academic predicts generative AI will gain more power than you might think – and maybe at your expense.

It’s early 2024 and the predictions are coming in fast and furious.

That’s especially the case with artificial intelligence and the primary underlying technologies powering AI such as large learning models and machine robotics.

One predictive entrant this week comes from Kentaro Toyama, a professor of community Information at the University of Michigan.

In comments to the news and opinion platform The Conversation this week, Toyama lays the groundwork for a 2024 AI experience that may buckle the knees of the most level-headed and rational C-suite executive.

“In 1970, Marvin Minsky, the AI pioneer and neural network skeptic said, “In from three to eight years we will have a machine with the general intelligence of an average human being,” Toyama notes. While the moment is not on society’s doorstep,” it’s safe to say that Minsky was off by at least a factor of 10,” Toyama says.

In 2024, Toyama says he expects more powerful artificial intelligence, in addition to a flood of new AI applications.

“The big technical question is how soon and how thoroughly AI engineers can address the current Achilles’ heel of deep learning,” Toyama says. “Will quick tweaks to existing neural-net algorithms be sufficient, or will it require a fundamentally different approach? Armies of AI scientists are working on this problem, so I expect some headway in 2024.”

Taking Over?

The human element factors into an even more dramatic prediction from Toyama – that AI may take more control over people’s lives than anyone seems to find comfortable.

“New AI applications are likely to result in new problems, too,” he says. “You might soon start hearing about AI chatbots and assistants talking to each other, having entire conversations on your behalf but behind your back.”

Some of that activity will “go haywire– comically, tragically, or both,” Toyama adds. “Deepfakes, AI-generated images and videos that are difficult to detect are likely to run rampant despite emerging regulation causing more sleazy harm to individuals and democracies everywhere.”

“And there are likely to be new classes of AI calamities that wouldn’t have been possible even five years ago,” he adds.

What’s needed quickly is more public and private oversight of AI and it’s not exactly barreling in on the horizon.

“Along those lines, what I most hope for 2024 – though it seems slow in coming – is stronger AI regulation, at national and international levels,” Toyama adds.


Recent Posts

Leave a Reply

Your email address will not be published. Required fields are marked *