@geoffreylitt@NotAlanKay Incredibly well-made argument!
The choice of when the cognitive debt is worth it vs when it isn't is just another trade-off like tech debt.
Fable is an amazing model but perhaps more importantly it taught us to be more ambitious in how we use agents even when they are operating with lesser models.
Why do all models love using "the smoking gun" while debugging?
Models definitely have peculiar lingo but it's interesting how they all converge to same peculiar lingo.