2 minute read

I love finding content that’s easy to share with people curious about AI but unsure where to start. Ethan Mollick’s post, Good Enough Prompting – Don’t Make This Hard, is perfect for anyone looking to dive in. It’s a quick read that cuts through the noise and offers actionable advice. I’ve pulled out some of the most useful points below, but I’d still encourage you to check out the full post.

His advice and style is extremely approachable. He cuts through the noise around “prompt engineering” and reminds us it’s not some magical skill – it’s mostly just clear communication. The best way to get started is to start. Dive in and put in the repetitions to learn how these tools work.

Here’s a summary:

  • Spend 10 hours using AI to understand it: The best way to learn AI is by using it on tasks that matter to you. Through practice, you’ll develop an intuitive sense of its capabilities and how to prompt effectively. Don’t aim for perfection – just experiment and learn as you go.

  • Treat AI like a capable but forgetful coworker: Think of AI as a highly skilled colleague who forgets everything after each conversation but is infinitely patient. Use it collaboratively to figure out its strengths and weaknesses. This mindset helps you avoid treating AI too narrowly, like a search engine or basic tool.

  • Start with familiar tasks to test its limits: Use AI in areas you’re already an expert in, so you can quickly assess if it’s accurate or plausible. Don’t let hallucinations discourage you – spotting them is easier in your domain, and you’ll learn when they matter and when they don’t.

  • Be specific and clear in your requests: AI performs best with detailed, specific instructions. Instead of asking for a general report, frame your request with context, audience, and purpose (e.g., “a report on remote learning tailored for a Midwestern university dean”). Clear directions dramatically improve output quality.

  • Provide examples and step-by-step instructions: Guide the AI by including examples of good responses or explicit steps for completing a task. For instance, ask it to generate a report in a specific structure or give feedback on its output, just as you would a human coworker.

  • Allow AI to admit uncertainty: Reduce hallucinations by encouraging AI to say, “I don’t have enough information to answer this.” This builds trust in its responses and helps you focus on areas where it’s most reliable.

  • Leverage its infinite patience to explore abundance: AI doesn’t get frustrated, so ask for multiple drafts, tones, or ideas to inspire your work. Push for variety (“make this 80% more creative”), combine outputs, or expand ideas to unlock breakthroughs. AI can generate thousands of options before repeating itself – use this abundance to your advantage and see where it takes you.

Mollick’s post is a great reminder that simplicity is key to using AI effectively. With curiosity, clear communication, and a willingness to iterate, you’ll quickly see the value these tools can bring to your work. Just dive in and give it a shot!

Updated: