A developer shares progress on a conversational symbolic AI architecture developed over the past eight months, aiming to address gaps in non-neural information extraction.
The author argues that symbolic architectures provide significant advantages in computational efficiency and the capacity to maintain extensive reasoning chains. The proposal involves extracting structured internet data via non-neural methods to feed the model, potentially enabling fluency comparable to large neural models. The approach relies on computational heuristics for reasoning tasks.
The developer is currently programming their ideas and seeks collaboration with other developers.