Hi Reader, Before today’s tip, I wanted to give you a heads up: Tomorrow, I’ll be launching something brand new! Watch out for the announcement 👀 👉 Tip #53: How to get great results from AI models through promptingIn the year after ChatGPT was released, I remember noticing two new trends:
My takeaway was that “prompting” must be really complex, and it was going to take a huge investment of time in order for me to get good at it. (In case you’re unfamiliar: “prompting” just means “inputting text into a Large Language Model.” Or more simply: “typing into ChatGPT.”) Since then, I’ve come to three conclusions about prompting:
📝 Simple tips for more effective promptsHere are my top tips for getting better at prompting, and thus increasing the chance that you will get great results from an LLM: Be very specific about what you want: How long should the answer be? What should the answer include? Who is the specific audience? What is the end goal of this task? Ask it to iterate: Within a given conversation, the model will remember everything you’ve said and everything it has replied (until it reaches the end of its “context window”). Thus if you’re not 100% satisfied with its answer, then tell it exactly how you want to improve! Give it a role: If you’re asking it a complex legal question (for example), you might get better results if you tell it to act like an expert lawyer. You can also use roles to adjust the writing tone of the results. Give it examples: If you need the answer in a specific format, give the model a few examples of what the output should look like for a given input. Ask it to think step-by-step: For complex tasks, guiding the model in how to approach the task step-by-step can result in better output. (The Claude documentation has a good example of this.) Try a different model: Every LLM is different, so if one isn’t helpful for a particular task, then just try another! I’ve talked previously about how to do this using Chatbot Arena or Typing Mind. 🧠Develop a mindset of abundanceI’ll leave you with one final tip from Ethan Mollick’s article, Getting started with AI: Good enough prompting: You don’t need to ask for one email, ask for three in different tones to inspire you. You don’t need to ask for one way to complete a sentence, ask for 15 options and see if that unlocks your writing. Don’t ask for 5 ideas, ask for 30. In fact, our research found that GPT-4 can generate thousands of ideas before a large percentage of them start to overlap. Your job becomes one of pushing for variation (“give me ideas that are 80% weirder”), recombination (“combine ideas 12 and 16”) and expansion (“more ideas like number 12”), before selecting one you like. 👋 See you on Wednesday!Be on the lookout for my launch announcement tomorrow! - Kevin P.S. Call your grandmother​ ​ Did someone AWESOME forward you this email? Sign up here to receive weekly Artificial Intelligence tips! |
Join 25,000+ intelligent readers and receive AI tips every Tuesday!
Hi Reader, Here are your top AI stories for the week: ChatGPT can weaken your brain Claude shares nerve gas recipe Amsterdam ends AI experiment due to bias Read more below! 👇 Sponsored by: Brain.fm Transform Your Focus With Brain.fm I know you're always on the hunt for tools that genuinely improve your life—which is why I'm excited to introduce you to Brain.fm's groundbreaking focus music. Brain.fm's patented audio technology was recently validated in a top neuroscience journal, showing how...
Hi Reader, Last week, I invited you to help me test Google's Data Science Agent in Colab, which promises to automate your data analysis. Does it live up to that promise? Let's find out! 👇 Sponsored by: Morning Brew Business news you’ll actually enjoy Join 4M+ professionals who start their day with Morning Brew—a free daily newsletter that makes business, tech, and finance news genuinely enjoyable to read and hard to forget. Each morning, it breaks down complex stories in plain English—cutting...
Hi Reader, Today I'm trying something brand new! I wrote short summaries of the 5 most important AI stories this week, and also turned it into a video: Watch the video I'd love to know what you think! đź’¬ AI-generated TV ad airs during NBA finals Prediction market Kalshi just aired this AI-generated ad on network TV during the NBA finals. It was created in just two days by one person using Google's new Veo 3 video model, plus scripting help from Google's Gemini chatbot. Expect to see many more...