5 ms·

This is coming from Dileep George and the Jeff Hawkins adjacent theories around hierarchical temporal memory, cortical columns, and much of the higher level the

This is coming from Dileep George and the Jeff Hawkins adjacent theories around hierarchical temporal memory, cortical columns, and much of the higher level theory around what it is that the brain is doing in its smallest, repeated functional units. They split some time back, but they're both quite rigorous and have done exciting research. This paper goes over allocentric framing of learning, allowing cortical networks to be built through learning and thinking over time, driving a more complex and nuanced model of the world to be built throughout the networks in the brain, where the hippocampus seems to hold the long term memory, with connections that feed back up to nearly every region in the cortex. It's not complete, but it abstracts a level away from phenomenological observations like the notion of Jennifer Aniston cells, or mirror neurons, or things that occur as a consequence of some underlying functionality. We're getting much closer to a complete picture of the algorithms underlying human intelligence, and those may unlock human level machine intelligence, the understanding of consciousness, and all sorts of great medicine and technology. Google made a good choice in hiring Dileep George.

K0SM0S 2y ago
Agreed. Anecdotal (but deep) research led me to postulate that our entire "inner world", for lack of a better word, is an emergent construction based on a fundamentally spatiotemporal encoding of the external world. This assumes that feeding and motility, i.e., a geometric interpretation of the external world, is among the first 'functions' of living organisms in the evolutionary order. They subsequently became foundational for neuronal systems when these appeared about 500 million years ago. The hypothesis was informed by language notably, where most things are defined in spatial terms and concepts (temporal too, though more rarely), as if physical experiences of the world were the building blocks of thinking, really. A "high" council, a "sub" culture, a "cover", an "adjacent" concept, a "bigger" love, a "convoluted" or "twisted" idea, etc. Representations in one's inner world are all about shape, position, and movement of things in some abstract space of sorts. This is exactly how I'd use a 4D modeling engine to express a more 'Turing-complete' language, a more comprehensive experience (beyond movement: senses, intuitions, emotions, thoughts, beliefs…): use its base elements as a generator set to express more complex objects through composition in larger and/or higher-dim space. Could nature, Evolution, have done just that? Iteratively as it conferred survival advantages to these genes? What would that look like for each layer of development of neuronal—and later centralized "brain"—systems? Think as in geometric algebra, maybe; e.g., think how the metric of a Clifford algebra may simply express valence or modality, for those neuronal patterns to trigger the proper neurotransmitters. In biological brains, we've already observed neural graphs up to 11 dimensions (with a bimodal distribution peak around ~2.5D and ~3.8D iirc… Interestingly for sure, right within the spatiotemporal ballpark, seeing as we experience the spatial world in 2.5D more than 3, unlike fishes or birds). Jeff Hawkins indeed strongly shaped my curiosity, notably in "A Thousand Brains" and subsequent interviews. The paper here immediately struck me as very salient to that part of my philosophical and ML research—so kinda not too surprised there's history there. And I'm really going off on a tangent here, but I'm pretty sure the "tokenization problem" (as expressed by e.g. Karpathy) may eventually be better solved using a spatiotemporal characterization of the world. Possibly much closer to real-life language in biological brains, for the above reasons. Video pretraining of truly multimodal models may constitute a breakthrough in that regard, perhaps to synthesize or identify the "ideal" text divisions, a better generator set for (any) language.
- IIAOPSW 2y ago
  I'm of the position this might be correct in the specific case of humans, but not fundamental to the algorithms of consciousness. Eg we could have similar emergent phenomena in algorithmic trading bots where all the emergent constructions are defined in terms of money and financial concepts rather than spatial concepts. They live in a reality of dollar signs rather than physical dimensions. That's neither inherently better nor worse. In fact, I'm somewhat of the position that nearly any grounding in a domain of shared objects where signalling is inexpensive would be suitable. That said, AI agents which grew up in some alien domain of shared objects would find us as unintuitive to reason about as we find quantum mechanics uninituitive to reason about. If the goal is AI that acts and talks like us, your way may be the way to go.
  - K0SM0S 2y ago
    I've no idea what the c-word means (consciousness), so I'll leave that aside; everything else checks out as absolutely sensible to me. Your last sentence strikes me as particularly validating. "My way", this framework, was meant to give a mechanistic description of our individual, subjective "inner world." Much like physics speaks of the outer, shared world; and in compliance with all objective 'hard' sciences. Indeed, it lends itself particularly well to be exploited by AI, notably in terms of architecture and domain-selection (by whatever core we call 'sapience') within a "Mixture-of-Experts" paradigm of sorts—which biology seems to have done: dedicated organs or sub-parts for each purpose, the Unix way to "Do one thing and/to do it well."
- sameoldtune 2y ago
  There’s an idea in psych that a high IQ correlates more than anything to an increased ability to navigate complex spaces. That’s what we do when we program, we create conceptual spaces and then imagine data flowing through them. And it is also why being intelligent in that way is seemingly so useful in everyday situations like budgeting, avoiding injury, and navigating institutions. It’s not all roses though—to quote Garrison Keillor, “being intelligent means you will find yourself stranded in more remote locations”
  - 2y ago
GenerWork 2y ago
>the notion of Jennifer Aniston cells I thought this was a joke that was slipped in, but it turns out this is actually a real thing[0]. [0]: https://en.wikipedia.org/wiki/Grandmother_cell https://en.wikipedia.org/wiki/Grandmother_cell
- MadnessASAP 2y ago
  Every so often I see an article or paper so far out of my field/familiarity that I'm not sure if I'm being pranked (i.e. The Retro Encabulator). This would be one of those papers. They could tell me they discovered synchronized Cardinal Grammeters in neurons and I would have no idea if they were joking.
  - ziofill 2y ago
    Agreed, but it's in a paper in Science, not a random blog post.
  - throwup238 2y ago
    The concept of a Jennifer Aniston neuron was originally introduced as a joke, a rhetorical foil for other theories.
    - IIAOPSW 2y ago
      A well known hypothetical cat was born out of a similar attempt at reductio ad absurdum.
      - namaria 2y ago
        Always a great signal people have no idea what they are talking about when one starts 'explaining' quantum weirdness using it.
- Mistletoe 2y ago
  > However, in that year UCLA neurosurgeons Itzhak Fried, mentee Rodrigo Quian Quiroga and others published findings on what they would come to call the "Jennifer Aniston neuron".[5][6] After operating on patients who experience epileptic seizures, the researchers showed photos of celebrities like Jennifer Aniston. The patients, who were fully conscious, often had a particular neuron fire, suggesting that the brain has Aniston-specific neurons.[6][7] This is fascinating. How many celebrity neurons does my brain contain? What is the limit? I used to read Perez Hilton, what happened to the neurons for C list celebrities I forgot? Were they repurposed or is a Hilary Duff neuron just waiting there all my life until needed?
amy-petrik-214 2y ago
I've always been of the hippocampus that it's short not long term memory in units of one day. Evidence that's the only CNS location that generates new cells on the regular (besides the nose). Timestamp cells produced within. Burns / fine tunes neocortex during sleep. and such. Also noting that such thing as "an electric artificial hippocampus" has existed for some time - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3395724/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3395724/

K0SM0S 2y ago

Agreed. Anecdotal (but deep) research led me to postulate that our entire "inner world", for lack of a better word, is an emergent construction based on a fundamentally spatiotemporal encoding of the external world. This assumes that feeding and motility, i.e., a geometric interpretation of the external world, is among the first 'functions' of living organisms in the evolutionary order. They subsequently became foundational for neuronal systems when these appeared about 500 million years ago. The hypothesis was informed by language notably, where most things are defined in spatial terms and concepts (temporal too, though more rarely), as if physical experiences of the world were the building blocks of thinking, really. A "high" council, a "sub" culture, a "cover", an "adjacent" concept, a "bigger" love, a "convoluted" or "twisted" idea, etc. Representations in one's inner world are all about shape, position, and movement of things in some abstract space of sorts. This is exactly how I'd use a 4D modeling engine to express a more 'Turing-complete' language, a more comprehensive experience (beyond movement: senses, intuitions, emotions, thoughts, beliefs…): use its base elements as a generator set to express more complex objects through composition in larger and/or higher-dim space. Could nature, Evolution, have done just that? Iteratively as it conferred survival advantages to these genes? What would that look like for each layer of development of neuronal—and later centralized "brain"—systems? Think as in geometric algebra, maybe; e.g., think how the metric of a Clifford algebra may simply express valence or modality, for those neuronal patterns to trigger the proper neurotransmitters. In biological brains, we've already observed neural graphs up to 11 dimensions (with a bimodal distribution peak around ~2.5D and ~3.8D iirc… Interestingly for sure, right within the spatiotemporal ballpark, seeing as we experience the spatial world in 2.5D more than 3, unlike fishes or birds). Jeff Hawkins indeed strongly shaped my curiosity, notably in "A Thousand Brains" and subsequent interviews. The paper here immediately struck me as very salient to that part of my philosophical and ML research—so kinda not too surprised there's history there. And I'm really going off on a tangent here, but I'm pretty sure the "tokenization problem" (as expressed by e.g. Karpathy) may eventually be better solved using a spatiotemporal characterization of the world. Possibly much closer to real-life language in biological brains, for the above reasons. Video pretraining of truly multimodal models may constitute a breakthrough in that regard, perhaps to synthesize or identify the "ideal" text divisions, a better generator set for (any) language.

IIAOPSW 2y ago
I'm of the position this might be correct in the specific case of humans, but not fundamental to the algorithms of consciousness. Eg we could have similar emergent phenomena in algorithmic trading bots where all the emergent constructions are defined in terms of money and financial concepts rather than spatial concepts. They live in a reality of dollar signs rather than physical dimensions. That's neither inherently better nor worse. In fact, I'm somewhat of the position that nearly any grounding in a domain of shared objects where signalling is inexpensive would be suitable. That said, AI agents which grew up in some alien domain of shared objects would find us as unintuitive to reason about as we find quantum mechanics uninituitive to reason about. If the goal is AI that acts and talks like us, your way may be the way to go.
- K0SM0S 2y ago
  I've no idea what the c-word means (consciousness), so I'll leave that aside; everything else checks out as absolutely sensible to me. Your last sentence strikes me as particularly validating. "My way", this framework, was meant to give a mechanistic description of our individual, subjective "inner world." Much like physics speaks of the outer, shared world; and in compliance with all objective 'hard' sciences. Indeed, it lends itself particularly well to be exploited by AI, notably in terms of architecture and domain-selection (by whatever core we call 'sapience') within a "Mixture-of-Experts" paradigm of sorts—which biology seems to have done: dedicated organs or sub-parts for each purpose, the Unix way to "Do one thing and/to do it well."
sameoldtune 2y ago
There’s an idea in psych that a high IQ correlates more than anything to an increased ability to navigate complex spaces. That’s what we do when we program, we create conceptual spaces and then imagine data flowing through them. And it is also why being intelligent in that way is seemingly so useful in everyday situations like budgeting, avoiding injury, and navigating institutions. It’s not all roses though—to quote Garrison Keillor, “being intelligent means you will find yourself stranded in more remote locations”
- 2y ago

GenerWork 2y ago

>the notion of Jennifer Aniston cells I thought this was a joke that was slipped in, but it turns out this is actually a real thing[0]. [0]: https://en.wikipedia.org/wiki/Grandmother_cell https://en.wikipedia.org/wiki/Grandmother_cell

MadnessASAP 2y ago
Every so often I see an article or paper so far out of my field/familiarity that I'm not sure if I'm being pranked (i.e. The Retro Encabulator). This would be one of those papers. They could tell me they discovered synchronized Cardinal Grammeters in neurons and I would have no idea if they were joking.
- ziofill 2y ago
  Agreed, but it's in a paper in Science, not a random blog post.
- throwup238 2y ago
  The concept of a Jennifer Aniston neuron was originally introduced as a joke, a rhetorical foil for other theories.
  - IIAOPSW 2y ago
    A well known hypothetical cat was born out of a similar attempt at reductio ad absurdum.
    - namaria 2y ago
      Always a great signal people have no idea what they are talking about when one starts 'explaining' quantum weirdness using it.
Mistletoe 2y ago
> However, in that year UCLA neurosurgeons Itzhak Fried, mentee Rodrigo Quian Quiroga and others published findings on what they would come to call the "Jennifer Aniston neuron".[5][6] After operating on patients who experience epileptic seizures, the researchers showed photos of celebrities like Jennifer Aniston. The patients, who were fully conscious, often had a particular neuron fire, suggesting that the brain has Aniston-specific neurons.[6][7] This is fascinating. How many celebrity neurons does my brain contain? What is the limit? I used to read Perez Hilton, what happened to the neurons for C list celebrities I forgot? Were they repurposed or is a Hilary Duff neuron just waiting there all my life until needed?

amy-petrik-214 2y ago

I've always been of the hippocampus that it's short not long term memory in units of one day. Evidence that's the only CNS location that generates new cells on the regular (besides the nose). Timestamp cells produced within. Burns / fine tunes neocortex during sleep. and such. Also noting that such thing as "an electric artificial hippocampus" has existed for some time - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3395724/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3395724/