This is more likely a result of neural networks being very general: General in the sense that they are able to approximate/model a lot of dynamic systems that allow operadic composition.
And the fact that we know that and now have proof in terms of intelligence of ChatGPT, is to me pretty clear that achieving AGI is possible if you crack the way to structure your neural network.
What.
It's definitely not human intelligence if that's what you are wondering.
What we may infer from that is that after a lot of time of evolution, functions depending on a lot of parameters appear. Yet Biological Neural Circuits seem quite bad at approximating simple continuous functions such as the multiplication between two numbers.
Gene Neural Network and Biological Neural Circuit are quite impressive structures considering their size, materials, and energy constraints. However if you allow bigger size, more energy, and faster conducting materials, it should be possible to do have faster modeling tools.
Moreover, to better take into account non-linear functions, my two cents is that Perceptrons could be further enhanced using piecewise polynomial functions instead of piecewise linear functions [4,5].
[1]: https://en.wikipedia.org/wiki/Gene_regulatory_network
[2]: https://en.wikipedia.org/wiki/Neural_circuit
[3]: https://en.wikipedia.org/wiki/Perceptron
That seems like almost fundamental emergent idea that is behind all the intelligence in the World.
It seems to allow for intelligence to occur, without this type of concept everything would just be random and chaotic.
The point is, how amazing it is that complicated behaviour can arise from this simple idea.
At this point it seems, that neural networks should really be taught at schools as soon as possible.
That is indeed quite nice, yet intelligence seems much more diverse in biological organisms and not always reduced to nodes and edge strengths. The nodes and edges are a way to store data and to modify it that seems particularly well adapted mechanisms based on electricity.
On the other hand, there exists other very different (and slower) biological intelligent mechanisms that are not based on electricity. For example, a tree is perfectly adapted to its environment, capable of taking energy from the sun, materials from the ground, transform it, and so on. Yet the intelligent mechanism that created trees is not based on neural networks as far as I understand it. In this mechanism, the data is stored in DNA, and DNA can be rewriten at each new generation. It is a completely different (and slower) approach.
Finally I agree that the emergence of fast intelligence through neural network is incredible, although for me the first impressive advance is the use of the electricity to speed up the information exchange. The arrangements in nodes and edges could follow naturally from the fact that the transmission of electric signals works better through 1D circuits. The second impressive advance is the way the networks are layered. This is critical and very difficult to have an arrangement of edges and nodes that can both be trained and inferred efficiently. Without the proper arrangement, a biological neural network or a computer circuit is much less intelligent.
Finally in the future we could imagine intelligent mechanisms based on quantum mechanics for example. In this case, it would probably be vastly different from nodes and edges, due to the underlying physical constraints that are different in electricity and in quantum mechanics.
tldr: yes nodes and edges seem fundamental in fast intelligence based on electricity, but other very different intelligent mechanism exist in biological organisms, and other very different intelligent mechanisms could still be invented in the future.
Consider that the uncertainty relations in physics fall neatly out of Fourier Analysis, which could suggest that the universe also does Fourier Analysis to figure out where things are and how fast they're going. But really, its that classical interaction is a sort of mechanical Fourier transform.
I think that moment where it feels like we brush against the real clockwork of the universe has an eldritch, cosmic-horror quality to it, which is both addictive and maddening.
Nothing profound at all about any of that. Just very general structures.
If DNA of some sequence, then interaction with electromagnetic fields, gravity, strong/weak nuclear force of xyz properties will probably result in mutation along path abc, a theory the speaker must provide replicable statistical analysis to back up or it goes in the bin.
IT obsession with data models has pushed IT workers into pseudoscience that ignores entropy, Lindy effects; structure is mutable. Physical forces are not. Data structures are not physical forces. They’re leaky mental models our biology garbage collects due to entropy and Lindy effects.
Anyway just the take of an EE who finds the obsession with software development bizarre, full of statistical mirages, hallucinated accomplishments.
Numerous strategies now exist to write cellular decision making programs using synthetic circuits. We are entering an era where we can write DNA programs and put them into people; I wonder how we can interest more CS-minded people in this kind of "synthetic biology as programming", especially as we move from proof of concept studies in bacteria to real trials in humans [2].
[1]: https://pubmed.ncbi.nlm.nih.gov/33243890/ [2]: https://arsenalbio.com/2023/05/16/arsenalbio-announces-prese...
I'm imagining a biological variant of eg. Verilog or VHDL, as this seems like the kind of domain where tools that allow for formal verification would be highly desirable.
Also, there is a major deficit of software engineering talent in biology research (probably because pro SWEs are too expensive for academic labs, and those labs are where much of the foundational research is done). If you have the bandwidth, part time / volunteer work with an academic lab in your area could be a great way in the door.
The GRNs are a bit more dynamic than strictly on/off, but the fact they have multiple types of interaction, such as enhancers and inhibitors, means they can model logical gates better than I understand a typical neural network can.
> Some proteins though serve only to activate other genes, and these are the transcription factors that are the main players in regulatory networks or cascades. By binding to the promoter region at the start of other genes they turn them on, initiating the production of another protein, and so on. Some transcription factors are inhibitory. [1]
These networks are similar to neural networks in that they process information through a series of interconnected nodes (genes and proteins, in the case of GRNs) that influence each other's activity. The activation or inhibition of one gene by a transcription factor can trigger a cascade of events, much like the way neurons activate each other in a neural networks.
[1] https://en.m.wikipedia.org/wiki/Gene_regulatory_network