Incoming links: Artificial neural network
Deep learning is the name artificial neural networks basically converged to in the 2010s/2020s.
It is a bit of an unfortunate as it suggests something like "deep understanding" and even reminds one of AGI, which it almost certainly will not attain on its own. But at least it sounds good.
It is a bit hard to decide if those people are serious or not. Sometimes it feels scammy, but sometimes it feels fun and right!
Particularly concerning is the fact that they are not a not-for-profit entity, and it is hard to understand how they might make money.
Charles Simon, the founder, is pretty focused in how natural neurons work vs artificial neural network models. He has some good explanations of that, and one major focus of the project is their semi open source spiking neuron simulator BrainSimII. While Ciro Santilli believes that there might be insight in that, he also has doubts if certain modules of the brain wouldn't be more suitable coded directly in regular programming languages with greater ease and performance.
FutureAI appears to be Charles' retirement for fun project, he is likely independently wealthy. Well done.
- www.aitimejournal.com/interview-with-charles-simon-ceo-and-founder-futureai
- 2022 raised 2 million USD:
Creativity and AGI by Charles Simon's at AGI-22 (2022)
Source. Sounds OK!- youtu.be/ivbGbSx0K8k?t=856 general structure of the human brain 86B total, matching number of neurons in the human brain, with:
- 14B: brainstem
- 16B: neocortex
- 56B: cerebelum
- www.youtube.com/watch?t=1433 some sequencing ideas/conjectures
Machine Learning Is Not Like Your Brain by Future AI (2022)
Source. Contains some BrainSimII demos.Ciro Santilli has felt that perhaps what is missing in 2020's AGI research is:
- the interface between:The key question is somewhat how to extract symbols out of the space-time continuous experiences.
- the continuous/noisy level (now well developed under artificial neural network techniques of the 2010's)
- and symbolic AI level AI
- more specialized accelerators that somehow interface with more generic artificial neural networks. Notably some kind of speialized processing of spacial elements is obviously hardcoded into the brain, see e.g. Section "Grid cell"
Forcing these boundaries to be tested was one of the main design goals of Ciro's 2D reinforcement learning games.
In those games, for example:Therefore, those continuous objects would also have "magic" effects that could not be explained by "simple" "what is touching what" ideas.
- when you press a button here, a door opens somewhere far away
- when you touch certain types of objects, a chemical reaction may happen, but not other types of objects
Bibliography: