... then you would miss the core concept of Kahneman's thesis.
His book is very thorough in defining these systems, how they interact, and what situations each are a good fit for.
Fast thinking is important for low cost, instantaneous decisions. It's important so that your type 2 system isn't engaged all day, because that system consumes more energy and takes more effort to kick in.
The problem is that erroneous heuristics are frequently encoded in type 1. That doesn't mean it's useless -- there is an intuitive expertise that can be developed by correctly providing feedback to your type 1.
Try not to think of it like "good" versus "bad" thinking. Type I is incredibly powerful, but has all these gotchas.
The way I think of it is type 1 is like an associative or probabilistic cache. Lookups aren't always perfect -- sometimes type 1 answers an easier question rather than the one it asked -- but the idea is that you wouldn't be able to function if every single decision or interpretation had to be run through type 2.
In other words, type 1 is a shortcut to prevent your type 2 from being overloaded.
