Thanks for this thoughtful piece, Jason. Could you provide an example of the "constructionist format" of foundational model development referenced below? I want to better understand what you mean by domain-specific models being expanded to other domains.
"Practical approaches to developing foundational models in complex systems should be considered more in a constructionist format – where useful models are developed in specific domains and expanded to others."
Thank you for the inquiry. What I mean is that the construction of what might be considered foundational models are likely be developed over time via the synthesis and transfer learning between many task specific models and data sets vs attempting to boil the ocean of data.
Thanks for this thoughtful piece, Jason. Could you provide an example of the "constructionist format" of foundational model development referenced below? I want to better understand what you mean by domain-specific models being expanded to other domains.
"Practical approaches to developing foundational models in complex systems should be considered more in a constructionist format – where useful models are developed in specific domains and expanded to others."
Thank you for the inquiry. What I mean is that the construction of what might be considered foundational models are likely be developed over time via the synthesis and transfer learning between many task specific models and data sets vs attempting to boil the ocean of data.