Yeah agreed, all thinking is "fast" as it were, although with self prompting plus tool use one gets closer to slow thinking that results in more rational reason

Yeah agreed, all thinking is "fast" as it were, although with self prompting plus tool use one gets closer to slow thinking that results in more rational reasoning (e.g. get the data from wolfram alpha, plug it into a calculator api, return results). No guarantee that it'll be rational though, much like in humans. If it had absolutely no model of the world it would be unable to dynamically reason about it at all which it very much does [citation needed I guess], so there's definitely something there.