Renowned computer scientist and robotics researcher Rodney Allen Brooks is pushing back on all of the excitement about artificial intelligence, or AI, saying experts and the public have been overestimating its abilities.
In a comprehensive interview withIEEE Spectrum, Brooks — who has been called an expert on AI — says OpenAI's language models, that ChatGPT are based on, are not yet sufficient.
The MIT professor says the AI tools available to us now are far away from being sentient or smarter than human intelligence.
In a 2017 article titled "The Seven Deadly Sins," Brooks wrote that he wanted artificial intelligence to exist and spoke of AI in a positive light. But, he also wants the public to realize that true capabilities of current AI tech.
Brooks said, "Mistaken predictions lead to fear of things that are not going to happen. Why are people making mistakes in predictions about Artificial Intelligence and robotics?"
He said it is the predictions that cause him to have to "spend time pushing back on them."
Brooks told IEEE Spectrum, "The large language models are a little surprising. I’ll give you that. And I think what they say, interestingly, is how much of our language is very much rote, rather than generated directly — because it can be collapsed down to this set of parameters."
Rote describes the condition of something behaving in a mechanical or repetitive way in order to be learned.
"But in that 'Seven Deadly Sins' article, I said that one of the deadly sins was how we humans mistake performance for competence," Brooks said.
He believes the AI technology available now has large-scale limitations and was created to sound like a human, but not be able to reason like one.
"We don’t get that same level of competence from the performance of a large language model," Brooks said.
He says that he is "doing something very arcane" it gives him results.
"But what I keep having to do, and I keep making this mistake - it answers with such confidence, any question I ask. [AI] gives an answer with complete confidence, and I sort of believe it. And half the time, it’s completely wrong," he said.
"I spend 2 or 3 hours using that hint, and then I say, 'That didn’t work,' and it just does this other thing. Now, that’s not the same as intelligence. It’s not the same as interacting. It’s looking it up," Brooks said.
In 2017, Brooks said his motivation for working in robotics and AI has been to create autonomous agents "that operate much like beings in the world."
But in 'Seven Deadly Sins' he writes that AI has not seen that level of success in recent years.
Brooks, an Australian computer scientist and cofounder of iRobot, which produces robots for industries, the military and home use, said we'll "probably not" see much come out of the billions that has been reportedly invested in ChatGPT.
Brooks appeared point to AI's potential to spread misinformation, saying, "So you’ve got to box things in because it’s not a database. It just makes up stuff that sounds good. But if you box it in, you can get really much better language than we’ve had before."
In a Metaverse Post analysis, Cindy Tan writes that GPT-4 was not "fully reliable" and said it has the ability to "hallucinate facts" and "make reasoning errors."
Brooks also criticized AI's computational cost, calling the technology "enormously expensive in terms of the computing needed."
Brooks believes that AI and automation still have a long way to go. He mentioned Amazon's market disruption by using robotics in warehouses. He says 80 percent of warehouses in the U.S. have no automation.
But, Brooks says fears of AI and automation taking away entire career specialties in the near future are hyperbole.
"I was reviewing a government report a few weeks ago, and it said, 'Lawyers are going to disappear in 10 years.' So I tracked it down and it was one barrister in England, who knew nothing about AI," Brooks said.
"'Surely, if it’s this good, it’s going to get so much better that we’ll be out of jobs in 10 years.' There’s a lot of disaster hype. Someone suggests something and it gets amplified," he said.
In Real Life: Voices of Nature
Can artificial intelligence really help us talk to the animals?
