Day 3
AI Basics When to Use AI vs. When Not To
The 80% rule: use AI when the task is language-heavy and a good-enough draft saves you time. Avoid it when accuracy is critical and hard to verify.
BrainGem · braingem.ai/learn
Alright, here's the cheat code — the one framework that'll save you from the two biggest AI mistakes. Mistake one: not using AI when you should (wasting hours on stuff it handles in seconds). Mistake two: trusting AI when you shouldn't (sending something to a client that AI confidently got wrong). Both are avoidable once you know the sweet spot.
Here it is: use AI when the task is language-heavy and being 80% right is a useful starting point. Drafting emails, brainstorming ideas, summarizing long documents, reformatting data, explaining confusing things in plain English — these are all in the sweet spot. You get a solid first draft in 30 seconds, then you spend 5 minutes polishing. You just saved yourself 30 minutes of staring at a blank page. On the flip side, don't use AI (or verify very carefully) when accuracy is critical and hard to check. Legal advice, financial calculations, medical information, anything where a confident-sounding wrong answer has real consequences. AI doesn't know what it doesn't know — it'll give you a wrong answer with the same confidence as a right one.
Best mental model: AI is a very fast, very eager intern with incredible language skills and absolutely no common sense. Would you let that intern draft your emails? Absolutely. Would you let them sign your contracts? Absolutely not. Same rules apply.
💡 Try This Today
Think about one task you did this week that took more than 30 minutes. Ask yourself two questions: Was it mostly about working with language (writing, reading, summarizing, organizing)? And would an 80%-good first draft have saved you time? If yes to both — that's your first AI candidate. Try it with AI next time and see what happens.