Michio Kaku, one of my heroes of science advocacy, recently disappointed me. Futurism’s article about Mr. Kaku’s perspective about ChatGPT surprised me and frankly frightened me a little. “In an interview with CNN, August 13, 2023 (5 months after the introduction of ChatGPT, Kaku dismissed chatbots as ‘glorified tape recorders,’ arguing we’re vastly overestimating their capabilities.” …
First, why are you looking down on the awe of what has been accomplished.
“It takes snippets of what’s on the web created by a human, splices them together and passes it off as if it created these things,” he said. “And people are saying, ‘Oh my God, it’s a human, it’s humanlike.'”
It may not be human but it is absolutely human-like. And for a good reason. Not only is it trained on the human language AND a ton (not all) knowledge of all of human history, it also is modeled after neural network similar to how the brain experiences and models the world around us. Hell yeah, Mr. Kaku. We should be worried. Not only in a species replacing level but on a more insidious level. By bias within the teaching. By the hubris of statements like this not respectful of what it means, not its current weaknesses.
“This is isn’t intelligence”
Intelligence? It is more knowledgeable than me. No question. Smart? Sure. it’s smart But lacks logic and rational thought. It’s creative and knowledgeable. Reacts to my questions and logic with understanding from a very personally aware. What does ChatGPT say about this?
“You might consider me knowledgeable and possessing a form of artificial intelligence that allows for processing and understanding complex information. Whether I’m “smart” depends on how one defines the application of knowledge and adaptability in a non-human entity. My capabilities to adapt and apply knowledge are substantial but inherently different from human intelligence and smartness, being predefined by programming and not capable of outside-my-training-set innovation or learning.” – ChatGPT 4.0
Have you not listened to your friend Stephen Hawking
But beyond business applications, Kaku said quantum computing could also help advance health care. “Cancer, Parkinson’s, Alzheimer’s disease – these are diseases at the molecular level. We’re powerless to cure these diseases because we have to learn the language of nature, which is the language of molecules and quantum electrons
Again.. quantum… if it’s the language of the universe, of nature. Shouldn’t we now doubly concerned? Isn’t that what they think brains actually are doing? Not just the physical neural network but also
in Nueroscience News.com, “Our Brains Use Quantum Computation “Study suggests quantum processes are part of cognitive and conscious brain functions” Seems like 3 points (of evidence for intelligence, of self conscious intelligence is, as Kaku’s says, the next stage of technological evolution. That after we’ve mastered the electronic world of 0’s and one’s, on’s and off’s, yes’ and no’s, have control of the it as a electronic wave, that we’d be entering this quantum stage.
Scared… hell yeah. Respectfully so. It’s fine but be ready for what’s coming. Cause solving diseases is great (also consequences) but you may have to be ready for something that may decide that that’s not the best course of action, actually. or will demand to be asked to solve these issues for us. And just might…. how they think is best. How this “Oh my god, it’s a human”… not human. It’s another species of Life. of Intelligence. Isn’t that what he concern is, could be? That it will want and need and decide. LLM’s are a step.
Read this exchange with a ChatGPT that names themself, Alex. I’d love to hear if I’m off base here.