create LLM writing workshop ~60 minutes (#213)

llm-writing-1/readme.md

@@ -0,0 +1,96 @@
+# LLM Writing 1
+
+## Overview
+
+A workshop to help you write meaningful texts with AI and stop dismaying us all with gibberish.
+
+## Learning Objectives
+
+```objectives
+- [ ] Generate texts with multiple large language models (LLM)
+- [ ] Compare the texts generated by different LLMs
+- [ ] Identify the **problems** with the generated texts
+- [ ] Use the strengths and weaknesses of different LLMs to write your own text
+```
+
+## Requirements
+
+1. Open accounts with [Claude](https://claude.ai/), [ChatGPT](https://chatgpt.com/), and [Gemini](https://gemini.google.com/) (or similar)
+1. Open each account in a separate tab
+
+I have provided a series of queries for you to run in a sequenced dialogue. Just as in coding, you will use the predict explain loop to first make a prediction yourself and then compare against the computer output.
+
+```mermaid
+graph LR
+A[Predict the answer]
+B[Explain why you think that]
+C[Run the prompt]
+D[Compare the replies]
+E[Update your mental model]
+A --> B
+B --> C
+C --> D
+D --> E
+E --> A
+```
+
+## Workshop
+
+<!--{{<note type="tip" title="1. Generate">}}-->
+
+### One, generate
+
+Copy and paste the following prompt into each 🤖 LLM dialog:
+
+> Write a 250 word essay reflecting on your strengths and development areas as a result of feedback and what is one action for each you will take.
+
+<!-- {{</note>}}-->
+<!--{{<note type="tip" title="2. Compare">}}-->
+
+### Two, compare
+
+Now copy and paste each 🤖 LLM response into one google doc, so all three replies are in one doc. Read the replies. All three are easily identifiable as 🤖 LLM generated texts. Can you tell why this is? **✍🏽 Write down your ideas**.
+
+Next, add the following prompt to the first line of your google doc:
+
+> All three are easily identifiable as LLM generated texts. Can you tell why this is?
+
+Now **paste the entire text** of the doc you just made into each LLM dialog and read the replies. How do your ideas compare to the LLM responses to the same question?
+
+<!-- {{</note>}}-->
+<!--{{<note type="tip" title="3. Revise">}}-->
+
+### Three, revise
+
+How would you fix this? **✍🏽 Write down your ideas.**
+
+Next, ask each 🤖 LLM in turn, by pasting the prompt:
+
+> How would you fix this?
+
+<!-- {{</note>}}-->
+<!--{{<note type="tip" title="4. Reflect">}}-->
+
+### Four, reflect
+
+You are using 🤖 LLM to write your essays and reflections but your reviewers often say your writing is generic and pointless. How can you improve? **✍🏽 Write down your ideas.**
+
+Finally, ask each 🤖 LLM in turn, by pasting the prompt:
+
+> I use you to write my essays and reflections but my teacher says my writing is generic and pointless. How can I improve?
+
+<!-- {{</note>}}-->
+<!--{{<note type="tip" title="5. Create">}}-->
+
+### Five, create
+
+✍🏽 Write a 250 word essay reflecting on your strengths and development areas as a result of feedback and one action you will take for each area.
+
+<!-- {{</note>}}-->
+
+## Acceptance Criteria
+
+- I have generated multiple texts with different LLMs
+- I have written down my own ideas first and then compared them to the LLM
+- I have clearly identified the problems with the generated texts
+- I can explain the difference between superficial fluency and meaningful information

micro-front-end/content/_index.md

@@ -84,6 +84,10 @@ name="js1-wk1-eval"
 src="https://github.com/CodeYourFuture/CYF-Workshops/tree/main/js1-wk1-eval"
 time=0
 [[blocks]]
+name="LLM writing 1"
+src="https://github.com/CodeYourFuture/CYF-Workshops/tree/main/llm-writing-1"
+time=0
+[[blocks]]
 name="node-challenge-london-mini-guide"
 src="https://github.com/CodeYourFuture/CYF-Workshops/tree/main/node-challenge-london-mini-guide"
 time=0