Considering the amount of people making up s** here by accident or on purpose, this would demand some very sophisticated level of cautiousness and plausibility-check, as also being able to follow the discussion to interweave corrections. This would demand a level of attention-detail and an understanding of human interaction at least on the level of a proper human historian IMHO.
But wouldnt it be better to focus the "attention to detail" on the details - and let the AI handle the fluff explainations for such - example would be to restrict the model to only research certain aspects of the prompt's intent...
So get a listing of languages, companies, technologies, industries, resources, supply chains - that can be directly tied to the evolution of computing - and let it web them together?
