Our brain is a beautiful sea of unknown, the most advanced machine ever, and AI, as artificial intelligence, would never be able to be as intelligent as a newborn, let alone a PhD holder! We would never be able to run as fast as a cheetah. We can drive or fly faster than a cheetah, but we cannot RUN as fast as a cheetah.
When people like Mira Murati or Sam Altman claim they will build PhD-level intelligence even if they have built an AI with an IQ of over 100, that’s a misleading claim.
“If you look at the trajectory of improvement, systems like GPT-3 were maybe toddler-level intelligence… and then systems like GPT-4 are more like smart high-schooler intelligence. And then, in the next couple of years, we’re looking at PhD intelligence…” ~ Mira Murati (video in comment)
In the realm of intelligence, there are many verticals like someone is good in math and someone in literature. Someone is left-brained and someone is right-brained. We have at least a dozen types of intelligence. On top of intelligence, we have consciousness which would fuel the intelligence, subconsciously. Our intelligence is an unknown N-dimensional space of our intelligence, which we, the neuroscientists of the world, have only grasped the very surface of this sea of unknowns. When I was doing my PhD in computational neuroscience, all we were trying to understand was to create the computational model of the smallest neuronal network with half a dozen neurons, and the behavior of this small neuronal network is so complex, so beautiful, with resonance, it’s like a group dance.
While everyone is chasing for the bigger, go for the small.
We’re still very far from creating something as intelligent as a human brain. Murati and Altmanists claims are failing to understand intelligence, biology, and neural networks. The neurons in biology and AI are fundamentally different. These unsubstantiated claims are only for fueling an arms race in the world of AI. A race that wouldn’t have any positive impact for any.
