It’s highly likely that all the basic brain function of all lower animals amounts to laboriously evolved ‘graph stuff’. The problem I have, confirmed by my own experiments, is that evolving ‘graph stuff’ is way too slow to be useful to us. If there is a path to C elegans-level intelligence this way it could take centuries or millennia to evolve, even at supercomputer speeds.
But the cortex is not ‘graph stuff’. The cortex and what HTM attempts to model is a column of neurons with its own characteristic algorithms and data structures. The graph is fully connected but the data (SDR) relates to which synapses are active. A more intelligent brain has more columns.
The HTM bet is that software replicating cortical columns will give us AGI, even if we never figure out the lower levels. Which is why the place to start is mammals or birds.
I’m not making myself clear enough. I’m not suggesting that a graph is used to wire up individual cells.
I think more macro. The HTM block is one of the potential blocks in the graph. Maybe it’ll take time, but hey, what else is there? Right now people keep stacking up point neurons or spiking neurons in the hope that somehow they’ll randomly spawn into intelligence, the demands of doing so inspiring awesome hardware like Cerebras.
Re-appropriate some of that resource for some evolutionary graph algorithm simulation and fitness evaluation.
I have good news on this front, Paul Cisek just published his latest article on the evolution of the brain.
With it we can studying how animals evolved instead of simulating evolution.
Only if those columns are not attached to sensory inputs form the skin, otherwise elephants and whales would be discussing AGI on this forum instead of us…lol… yeah, I know the columns are different… just thought it was funny to take a comment way out of cortex, lol.
I think we just need something solid. Compare it to coding (especially things like optimization and multithreading) or scientific articles outside functional neuroscience. Those constantly deal with super subtle problems. We need to get to the point where that process can start in regards to how the brain does useful stuff.
Evidence suggests that cortical function is necessary but not sufficient for ‘higher’ functions such as language, sequencing, cooperation, tool-making, consciousness. Elephants, whales, crows and apes all have the cortex for it, but only a bipedal primate was able to take that next step. Why?
And it was relatively quick: cortex goes back well over 50 M years, but humans no more than around 150K. Software updates anyone?
The logic gates made of semiconductors in a tiny calculator, are essentially the same kind as those in a powerful computer that can play virtual reality games.
Maybe the human brain has similar cortical columns, just a lot more of them, and as a consequence it’s able to cram in a LOT more knowledge than an animal brain can.
It is incredible that a modern human has to go through so many years of schooling - think of a typical Ph.D. candidate - to engage in some productive research/academic work.
Consciousness ➔ introspection/metal time travel/internal visualization of oneself in a simulation/personal narrative over time.
The above requires language, which is necessary for consciousness but not sufficient. A cortex is also necessary but not sufficient
Consciousness is learned and is the operating system of the brain. It starts to emerge right around 3 years old and requires years of training to reach full maturity in the 20’s, which explains the brainless behavior of college students.