If the goal is maximizing skill at writing, one should use LLMs a lot. What you wrote about likely failure modes of doing so is true, but not an inevitable outcome. If Language Models are useful tools for writing, avoiding their use due to concerns about being unable to handle them is a mistake regardless of whether these concerns are warranted. Why?
if you’re trying to make a splash with your writing, you need to meet a much higher bar than the average person
Having aptitude necessary to “make a splash” is very rare. Not taking chances probably means one won’t reach the top. Especially if competent LLM use raises the ceiling of human capability.
“cyborgs”, a specific kind of human-in-the-loop system which enhances and extends a human operator’s cognitive abilities without relying on outsourcing work to autonomous agents
If the goal is maximizing skill at writing, one should use LLMs a lot. What you wrote about likely failure modes of doing so is true, but not an inevitable outcome. If Language Models are useful tools for writing, avoiding their use due to concerns about being unable to handle them is a mistake regardless of whether these concerns are warranted. Why?
Having aptitude necessary to “make a splash” is very rare. Not taking chances probably means one won’t reach the top. Especially if competent LLM use raises the ceiling of human capability.
Note that by competent use I mean something like cyborgism: https://www.lesswrong.com/posts/bxt7uCiHam4QXrQAA/cyborgism