Straight answers.

No. The tools change every few months, which means everyone is perpetually a beginner at the current version. Someone starting today is roughly one weekend behind someone who started a year ago, because almost none of what they learned was durable.

Not at first. The free tier of any major assistant is more than enough for your first month or two. Paid tiers mainly buy higher usage limits and somewhat stronger reasoning, which start to matter only once you have a habit worth protecting.

Any of the well-known ones. At the level of everyday use the differences are small and the comparison arguments are mostly sport. Pick one, use it for a month, and ignore the leaderboards.

Almost always because the request had no context. Assistants default to the most average possible answer when they do not know the purpose, the audience, or the constraints. Two extra sentences of context usually improves the result more than switching tools would.

Yes, and this is the single most important thing to understand. It will occasionally state something false in exactly the same confident tone it states something true, with no tell. Use it for things you can verify, or where being roughly right is fine.

Treat it like a work email rather than a diary. Turn off training on your conversations in settings, strip names and account numbers, and never paste someone else's private information. Most everyday use carries very little risk.

It does not replace taste, judgment, or having something to say, which is most of what creative work actually is. It is genuinely useful for the labour surrounding the craft — admin, drafts, variations, research. The concerns are legitimate; the practical answer is to use it where the craft isn't.

No. The prompt-template industry is largely theatre. Explaining what you want the way you would to a competent new assistant — what it is for, who it is for, what would be wrong — outperforms nearly every template on sale.

Something tedious that you already know the right answer to. Choosing a task you can judge means you learn instantly whether the result is good, which teaches you far more than an unfamiliar task would.

If after a month you are still deciding each time whether to use it, it has not landed yet. When it works, it becomes boring — you stop noticing you are using it, the way you stopped noticing you were using a calculator.