CLEVELAND — As artificial intelligence models grow at breakneck speed, there is a lot of uncertainty about what the future of the technology will look like.
That’s something that those who are building it are still figuring out.
“So, like the human body, we don’t really understand how things work,” said Vipin Chaudhary, a computer and data science professor at Case Western Reserve University. “So, the scientists try to figure out how things are working. Think of it as a similar case with the AI models: they work really well. We don’t know how they work.”
He said research into AI is not keeping pace with the release of new models, making it hard to predict what will happen next, and sometimes AI acts in ways that Chaudhary said researchers can’t yet explain.
“There was news about some of the open AI models on their own going out and hacking another big AI company on their own,” Chaudhary said.
There is a lot of fear about how AI could be misused, which is why Chaudhary believes there needs to be more regulation in place to control it, but doing that for a technology that is changing so rapidly presents challenges.
“This is moving very fast,” he said. “We have to move faster towards those regulations. It’s not also very clear what the regulations need to be.”
While not fully understanding how and why AI models reach the conclusions they do can raise concerns, it’s not slowing innovation.
“What we will get to next is this intelligence being embodied into robots,” Chaudhary said. “And the robots become intelligent as humans and better than humans. So, they will start doing a lot of these tasks. So, that does scare a lot of people.”
He imagines it could look like having a robot to help with chores around the house, teach children, or accompany people who are lonely.
For now, he said, he can’t say exactly how the technology will be used moving forward. It will take time to see how the research, social perceptions and regulations play out.