With the way Artificial Intelligence evolves, those that have access to stronger AI models, have advantage over their competitors, but there is another, deeply human, area that makes a difference. It's the ability to ask great questions.
Professional spellcheckers require quite a lot of trust. You basically let an external company read your emails, access your blog posts before they are published, and process documents with company secrets. But do you really need them when LLMs that you can run on your laptop match the quality of professional spellcheckers?
LLM's are just human-imitation machines. They don't learn from interactions, they are not reliable experts, and they don't care if you turn them off or not.