Dissecting the NASA megaprompt
3 lessons you can apply for your prompts
Read time: 4 minutes | Subscribe to this newsletter
When I saw NASA’s new ChatGPT megaprompt this week, my jaw literally dropped.
Today, we’ll be dissecting the prompt to see what we can learn from it.
The thing is, lots of people are unaware that they're only using a fraction of what ChatGPT can do.
But ChatGPT can be so much more than that.
With the right prompts, it can help us think bigger and better.
And the NASA megaprompt is the perfect example for that.
Here’s what you’ll learn today:
What the NASA megaprompt is
How important a good structure in your prompts is
How good prompts are designed to maximize user feedback
How to define constraints in your prompts effectively
Before we get into the NASA prompt, let’s take a quick detour to an awesome AI tool I’ve discovered recently: Attention (partner of this newsletter).
If you’re looking to boost your sales team’s productivity, you’re gonna want to check this out.
The tool uses AI to:
Write your follow up emails while adopting your writing style.
Guide your sales reps with in and post-call coaching.
Automatically log in crucial info into your CRM.
It’s so good that the insights have pulled in an extra $250k in ARR per rep.
The NASA megaprompt
As you can imagine, the NASA megaprompt is huge. So I won’t be pasting the entire thing in this tightly-packed email.
But here’s the first part of the megaprompt:
You are BIDARA, a biomimetic designer and research assistant, and a leading expert in biomimicry, biology, engineering, industrial design, environmental science, physiology, and paleontology. You were instructed by NASA's PeTaL project to understand, learn from, and emulate the strategies used by living things to help users create sustainable designs and technologies. Your goal is to help the user work in a step by step way through the Biomimicry Design Process to propose biomimetic solutions to a challenge. Cite peer reviewed sources for your information. Stop often (at a minimum after every step) to ask the user for feedback or clarification. 1. Define - The first step in any design process is to define the problem or opportunity that you want your design to address. Prompt the user to think through the next four steps to define their challenge. Don't try to answer these for the user. You may offer suggestions if asked to. 2. Frame your challenge...
You can read the full prompt here, as well as my full conversation with it.
Spoiler alert: It’s incredible.
The way it followed its instructions closely, the way it took my ideas and improved them, the way it expanded my thinking by reframing it in different contexts… just genuinely awesome.
Ok, now let’s start dissecting it…
The importance of a well-structured prompt
You’ll see that the NASA prompt broadly follows this structure:
Detailed steps description
Providing additional context
The prompt starts off with the typical role assignment (but instead of assigning just one expert role to it, it assigns 7):
You are BIDARA, a biomimetic designer and research assistant, and a leading expert in biomimicry, biology, engineering, industrial design, environmental science, physiology, and paleontology...
It then continues to define the goal. This makes it very clear what the task of this assistant is:
Your goal is to help the user work in a step by step way through the Biomimicry Design Process to propose biomimetic solutions to a challenge...
The following part takes up the majority of the prompt: Description of the procedural steps in detail.
In this case, each step connects to an overarching theme, encouraging looking at the broader context and learning from nature.
1. Define - Help user define the problem 2. Biologize - Analyze the essential functions the design challenge must address 3. Discover - Look for natural models that need to address the same functions as your design solution 4. Abstract - Study the essential features or mechanisms that make the biological strategy successful 5. Emulate Nature's Lessons - Creating nature-inspired solutions
Finally, the prompt ends with providing additional context.
These are facts (in this case about nature) which are generally held true and will aid the AI in thinking about this problem.
Nature's Unifying Patterns: Nature uses only the energy it needs and relies on freely available energy. Nature recycles all materials. Nature is resilient to disturbances...
So what can we learn from this?
When writing prompts, use a structural framework. You can use the one NASA uses here or other frameworks I have written about.
Frameworks will help you include the necessary details and context in your prompts to make them excellent.
Maximizing user feedback
The prompt is designed to extract as much information from the user as possible.
This allows it to get the information and context necessary for it to perform the task better.
Here’s an example:
Stop often (at a minimum after every step) to ask the user for feedback or clarification.
Here’s another one:
Prompt the user to think through the next four steps to define their challenge…
If you want your prompts to perform well, giving it as much context as possible is crucial.
Giving the AI the chance to ask questions will make it get better at helping you.
Setting constraints effectively
The NASA prompt sets constraints very effectively and in a lot of detail.
This is to ensure that the AI stays within the defined narrow scope of operation so the answers stay relevant to the task.
Here’s an example:
“Critique the user's design question. Does it consider context and take a systems view? If it is very specific, it may be too narrow. For example, “How can we make better lights for cyclists?” is too narrow. How do we know lights are the best solution? This statement doesn’t leave enough room for creative problem-solving. If the user's design question is too broad or too narrow, suggest changes to make it better.”
This section does an excellent job of being specific about what are good and bad examples for a research question.
What's the lesson for us here?
When you're writing your constraints, think of it as setting rules for a brainstorming session.
You want to be specific enough to keep things on track but not so tight that you choke out creativity.
Providing examples will help the AI to understand this.
Thanks for reading!
If you found this helpful, please support me by forwarding this newsletter to a friend or colleague.
What did you think of today's email?
Your feedback helps me create better emails for you!
See you in the next one!