4 ms·
It's interesting I so often see the discussion of AI in the context of what a singular model can or can't do. This seems myopic. Our own brains have specializ
by kromem 4y ago
It's interesting I so often see the discussion of AI in the context of what a singular model can or can't do.
This seems myopic.
Our own brains have specialized areas. Someone who has a stroke impacting Broca's region isn't going to have a fun time trying to do what that region was specialized for.
And as a society, I don't think we'd have gotten very far if every individual was expected to perform every task.
Where generalization in AI seems like the holy grail is not in replacing more specialized models, but in orchestrating them.
As impressive as this effort at generalization is, unless the work here is going to translate into orchestration and management of other specialized models, I don't think this is going to have much market relevance unless the generalized approach can somehow (in that generalization) gain an advantage in outperforming specialized approaches.
- yarg 4y agoOn the topic of Broca's area, families with mutations to the FOXP2 gene are known to have trouble with vocalisations. https://pubmed.ncbi.nlm.nih.gov/16437554/ https://pubmed.ncbi.nlm.nih.gov/16437554/ There are a number of structures related to specific functionality - facial recognition is similarly hardwired. Problems with the fusiform face area lead to prosopagnosia, it'll also be the region of the brain responsible for the hollow face illusion.
- mdp2021 4y ago> Our own brains have specialized areas But Intelligence is one of those, and when people start calling naïvely calling "intelligence" its opposite "imprint", there is a problem. There are even the seeds of a social problem. > orchestrating ... specialized models Intelligence (ontology development) is one such specialized model which I have seen more faithfully and consciously attempted in classic AI as opposed to contemporary. Orchestration is crucial for the peripherals that allow the input/output. The study of "generalization" is none the less crucial, because we are to investigate the development of a system that can intrinsically, by its own nature, be exported to the largest number of relatable skills. Even with the highest specialization, up to physical, there is a matter of organicity: a hand is "specialized", yet a hook is not a hand.