are agents turning machine?
is that even a reasonable question? or is it off as if asking are carbs bad for me or not?
Sounds sensible but meaningless without context.
this is the context for are agents turning machine or not?
Can an LLM-based agent, given unbounded memory and tool access, compute anything a Turing machine can?
that's not interesting.
better question is, can agents do things turning machines can't?
even better, can our brains do things turing machines can't?
second question is unsettled. haha.
another interesting question, can turing machines do everything?
we know one problem they can't solve.
halting problem.