**I Am GPTed** *Theme music fades in – upbeat, quirky synth with a glitchy AI beep* Hey there, misfits and AI newbies, welcome to **I Am GPTed**, where I, Mal – the Misfit Master of AI, or just Mal for short – dish out practical tips on wrangling ChatGPT, Claude, Gemini, Grok, and whatever LLM the tech bros dream up next. No fluff, no hype, just stuff that actually works. Because let's face it, I'm still figuring this out too, and if I can hack it, so can you. Today, we're diving into prompts that don't suck. First up: **Chain of Thought prompting**. It's like telling your buddy to think out loud instead of blurting nonsense. Tech hype says it's magic; really, it's just making the AI show its homework so you spot the dumb mistakes. Before example: "How do I fix my leaky faucet?" AI spits back a vague list, and you're still flooded. After: "How do I fix a leaky faucet? Think step by step: diagnose the type of leak, tools needed, safety first, then steps." Boom – it walks you through shutoff valve, washer swap, like a plumber who's not charging $200 an hour. Try it; your pipes – and prompts – will thank you. Practical use case for us normies? **Grocery planning on a budget**. Don't just ask "Meal plan for a week." Say: "I'm a busy parent, $100 budget, two kids who hate veggies. Chain of thought: list cheap proteins, hide veggies creatively, total under $100." Suddenly, AI spits out taco nights with blended spinach no one notices. Saved my sanity last week – and yeah, I ate the tacos. Common beginner mistake? **Vague prompts**. I did this for months: "Write a blog post." Got garbage. Avoid it by being bossy with specifics – role, tone, length, examples. I admit, I once prompted Grok for "dating advice" like a desperate teen. It told me to "be myself." Duh. Now I say: "You're a sarcastic wingman. Give 5 texts for asking out a barista without sounding creepy." Way better. Build skills with this simple exercise: Pick a boring task, like "email your boss." Chain-of-thought it: "Step 1: State purpose. Step 2: Key facts. Step 3: Call to action. Draft as helpful assistant." Tweak the output. Do three a day; you'll prompt like a pro by Friday. Last tip: Evaluating AI junk? **Reverse engineer it**. Ask: "Critique this as a picky editor: strengths, weaknesses, fixes." Or rate it 1-10 on accuracy, creativity, usefulness. If it's meh, feed back: "Make it punchier, less wordy." Iterate till it's gold. No more settling for robo-blah. That's your toolkit, folks – practical, no PhD required. If the tech overlords say it's revolutionary, it's probably just common sense. Subscribe now so you don't miss me mocking the next AI fad. Thanks for listening! This has been a Quiet Please production – head to quietplease.ai for more. Catch you next time, misfits. *Outro music swells – fade to glitchy beep* (Word count: 498)For more check out https://www.quietperiodplease.com/and for some great deals go to https://amzn.to/4nidg0PThis content was created in partnership and with the help of Artificial Intelligence AI

Ver todos los episodios