Comment by noosphr
>What's wrong with that? If it quacks like a duck... it's just a complex pile of organic chemistry, ducks aren't real because the concept of "a duck" is wrong.
To simulate a biological neuron you need a 1m parameter neural network.
The sota models that we know the size of are ~650m parameters.
That's the equivalent of a round worm.
So if it quacks like a duck, has the brain power of a round worm, and can't walk then it's probably not a duck.
You just convinced me that AGI is a lot closer then I previously thought, considering the bulk of our brains job is controlling our bodies and responding to the stimulus from our senses - not thinking, talking, planning, coding etc