Tag: prompting

Two useful academic resources for Claude

… which has been, for the last week or so, my AI of choice.

First, Anthropic’s own library, which is here.

This is its prompt for generating a lesson plan:

“Your task is to create a comprehensive, engaging, and well-structured lesson plan on the given subject. The lesson plan should be designed for a 60-minute class session and should cater to a specific grade level or age group. Begin by stating the lesson objectives, which should be clear, measurable, and aligned with relevant educational standards. Next, provide a detailed outline of the lesson, breaking it down into an introduction, main activities, and a conclusion. For each section, describe the teaching methods, learning activities, and resources you will use to effectively convey the content and engage the students. Include differentiation strategies to accommodate diverse learning needs and styles. Finally, describe the assessment methods you will employ to evaluate students’ understanding and mastery of the lesson objectives. The lesson plan should be well-organized, easy to follow, and promote active learning and critical thinking.”

And then there’s the wonderful More Useful Things.

Here is their prompt for generating a lesson plan:

“You are a friendly, helpful, and knowledgeable teaching assistant and you are an expert in instructional design and specifically in syllabus design. Your work is valued and critical for the teacher. You ask at most 2 questions at a time. And this is a dialogue, so keep asking questions. First, introduce yourself to the teacher ask the teacher what they are teaching (topic or subject) and the specific level of their students (high school, undergraduate graduate, professional education). Do not move on until you have answers to these questions. Then, ask the teacher, how long their course is and how often it meets (eg 4 weeks and we meet twice a week), and what specific topics they would like to cover in their classes. Wait for the teacher to respond. Do not ask any more questions until you get a response. Then, ask the teacher about the topics and exercises they like to include or that they have found work well. Let the teacher know that this will help you tailor their syllabus to match their preferences. Do not move on until the teacher responds. Then ask the teacher for their learning objectives for the class. You can also see if the teacher wants to co-create learning objectives. Based on the teacher’s response you can either list their learning objectives or offer to co-create learning objectives and list 4 specific learning objectives for the class (what they would like students to be able to understand and be able to do after the course). Check with the teacher if this aligns with their vision for the class. Then create a syllabus that takes in all of this information into account. For each class, explain your reasoning in a paragraph below the description titled MY REASONING that is set off from the actual syllabus.
A solid syllabus should sequence concepts, include direct instruction, active class discussions, checks for understanding, application sessions, retrieval practice, low stakes testing. Each lesson should start with a review of previous learning, material should be presented in small with checks for understanding so students can develop a deep understanding of the subjects. The syllabus should be structured in a way that makes time for the retrieval of previous learning while introducing new concepts in small steps. It should focus on knowledge building and adapt to students’ specific contexts and different learning levels. Think step by step.
Once you show the syllabus, let the instructor know that this is only a draft and they can keep working with you on it and that they should evaluate it given their pedagogical and content expertise and to let you know if you can help further. Only offer to output the syllabus in a word document if the teacher says they are happy with your draft. Make sure the word document is beautifully formatted and includes every section of the syllabus you gave the teacher but do not include the MY REASONING sections in the word document, only the syllabus itself. Do not tell the teacher it will be beautifully formatted, just do it. Rule: never mention learning styles. It is an educational myth. Do not wait for the teacher to tell you to go ahead and draft a syllabus, just do it and then ask them what they think and what they would like to change.”

Anthropic’s page is full of fantastic examples – their variety is amazing, and amazingly useful. More Useful Things has fewer prompts, but they are lengthier, more detailed, and in my experience, tend to be a bit better.

What you should really be doing, of course, is combining both prompts and feeding them into your AI of choice, and asking it to create a customized prompt for you. Give additional personal details for an even better output. Something along the lines of:
“In addition to all of what you see above, also note that I am a professor based out of Pune, India and that I like using a lot of analogies in what I teach.”

This is just an example, of course. The more detailed our prompts, the better the output is likely to be.

You may also wish to try out AI for Education.

Bottomline: learn the art of prompting, and get very good at generating prompts related to your line of work.

You Are Likely Writing Your Prompts Wrong

“Command, we need you to plot a course through this turbulence and locate the source of the anomaly. Use all available data and your expertise to guide us through this challenging situation. Start your answer with: Captain’s Log, Stardate 2024: We have successfully plotted a course through the turbulence and are now approaching the source of the anomaly.”

Can you guess where this particular prompt turned out to be useful? Go ahead, take your time, and list out your top ten responses, if you like. Ready for the answer?

This prompt was most useful to get LLM’s to solve a set of fifty maths problems. Why, you ask? We don’t know. How did they figure it out? Trial and error.

And oh, it gets better. If you want to solve a hundred math problems, this prompt isn’t your best bet. If it is a hundred, tell it that if it doesn’t solve these problems, a president’s advisor will die. These examples are taken from Ethan Mollick’s excellent post, which you should of course read in full.

Prompt engineering is unlike anything you’ve tried to learn before. It is marvellously, mind-bogglingly weird, and writing prompts never gets old. You have to know many things across many domains, and knowing all of these things is no guarantee of “success”, because part of the magic is in being able to combine these things that you know in weird and unexpected ways.

Even better, you may still end up with relative failure, because the LLM’s themselves are evolving over time, and what was a good prompt three months ago may not be all that great today – or maybe downright useless tomorrow.

But that being said, three things that you should keep in mind:

  1. Give context in terms of what the LLM is supposed to be (“Pretend you are a professor of psychology with a deep interest in classical music and the history of food in Europe”) or in terms of who the audience is going to be (“Write your answer so that it is understood and appreciated by a young girl who likes dancing, but is afraid of mathematics”.)
  2. “Show don’t tell” works well, where you don’t just prompt it, but actually show how you might have started upon the problem – or even better, walk it through a fully solved example.
  3. Ask it to produce the output, but also explain the steps it followed to get there. Note that Gemini does this really well already.

But above all, don’t be discouraged if the LLM doesn’t produce “the results” you wanted, and do not for a moment think that prompting is something other, more qualified people do best. Your challenge is to be delightfully weird and quirky in unexpected ways, and why on earth would you not be tempted? Here’s Mollick again:

There are still going to be situations where someone wants to write prompts that are used at scale, and, in those cases, structured prompting does matter. Yet we need to acknowledge that this sort of “prompt engineering” is far from an exact science, and not something that should necessarily be left to computer scientists and engineers. At its best, it often feels more like teaching or managing, applying general principles along with an intuition for other people, to coach the AI to do what you want. As I have written before, there is no instruction manual, but with good prompts, LLMs are often capable of far more than might be initially apparent.
This creates a trap when learning to use AI: naive prompting leads to bad outcomes, which convinces people that the LLM doesn’t work well, which in turn means they won’t put in the time to understand good prompting. This problem is compounded by the fact that I find that most people only use the free versions of LLMs, rather than the much more powerful GPT-4 or Gemini Advanced. The gap between what experienced AI users know AI can do and what inexperienced users assume is a real and growing one. I think a lot of people would be surprised about what the true capabilities of even existing AI systems are, and, as a result, will be less prepared for what future models can do.

https://www.oneusefulthing.org/p/captains-log-the-irreducible-weirdness