Does anyone else see parallels between how entorhinal cortex and hippocampus are connected vs how the neocortex and thalamus are interconnected? Both seem to contain a sensory loop with dual spatial context and I wonder if they may be related through evolution somehow.
I am reminded of Kim Peek who claims to have memorized 3 shelves of books at the Utah public library but doesn't know how to tie his shoes. Where was that stored? Do we all have the same capability to store all of that information, or was Peek wired differently?
Naive question from an non-expert here: Has anyone thought of taking a multi-modal LLM-maybe with robotic embodiment for added cross training, training it on everything from Jeff Hawkins books and papers, all neurology papers and texts, all AI papers and texts, maybe throw in psychology/psychiatry literature and lectures and such, then ask it to come up with testable hypotheses in neuroscience and new approaches to brain based software or hardware? These neural nets may not be genuinely intelligent but their simulation of intelligence is getting powerful enough that they might be useful in making progress along the path to genuine artificial intelligence.
Hello Numenta. I want to create an electronic brain on the basis of which to subsequently develop a strong AI. I have developments that will allow already now to make microchips with very high computing power, specialized in the calculation of neural networks. There are also developments to create the first prototypes of self-learning neural arrays based on silicon microchips. Unfortunately, despite all the interest of the world community in the topic of AI, in reality it is very difficult to find like-minded people with whom it would be possible to implement this task. It is even more difficult to find investors who are ready to invest in the development and production of microchips for calculating the operation of neural networks and devices based on them - neural maps. Tell me, can you be interested in my developments?
Thank you, please keep sharing these videos! :)
Have you looked at Hinton’s new Forward-forward algorithm? I’m curious about its biological plausibility
"The Forward-Forward Algorithm: Some Preliminary
Investigations"
Thank you for sharing. Great way to end my work day
Does anyone else see parallels between how entorhinal cortex and hippocampus are connected vs how the neocortex and thalamus are interconnected? Both seem to contain a sensory loop with dual spatial context and I wonder if they may be related through evolution somehow.
I am reminded of Kim Peek who claims to have memorized 3 shelves of books at the Utah public library but doesn't know how to tie his shoes. Where was that stored? Do we all have the same capability to store all of that information, or was Peek wired differently?
Naive question from an non-expert here:
Has anyone thought of taking a multi-modal LLM-maybe with robotic embodiment for added cross training, training it on everything from Jeff Hawkins books and papers, all neurology papers and texts, all AI papers and texts, maybe throw in psychology/psychiatry literature and lectures and such, then ask it to come up with testable hypotheses in neuroscience and new approaches to brain based software or hardware?
These neural nets may not be genuinely intelligent but their simulation of intelligence is getting powerful enough that they might be useful in making progress along the path to genuine artificial intelligence.
Thank god y’all learned how to use separate sections/chapters feature finally
This video is very exciting as a lot of this information is very congruent with my intuition
No Backprop....ever. Period.
it's a pity that ChatGPT didn't come from you
Hello Numenta. I want to create an electronic brain on the basis of which to subsequently develop a strong AI. I have developments that will allow already now to make microchips with very high computing power, specialized in the calculation of neural networks. There are also developments to create the first prototypes of self-learning neural arrays based on silicon microchips. Unfortunately, despite all the interest of the world community in the topic of AI, in reality it is very difficult to find like-minded people with whom it would be possible to implement this task. It is even more difficult to find investors who are ready to invest in the development and production of microchips for calculating the operation of neural networks and devices based on them - neural maps. Tell me, can you be interested in my developments?