Hey - I just want to thank you for reading this and for supporting all my work over the last 11 months. 

As a teacher myself, I genuinely try to offer value to teachers and educational leaders with AI. I get to see AI transforming the practice of teachers and offering amazing solutions for education. I will continue to work to help you and try to make a better education system for our young people.

You are my priority. 

In this edition:

• A new prompting framework to help you get the most out of AI
• UK and US government guidance about AI in education
• Making learning more accessible with AI from Google & Dan Fitzpatrick

It's been a privilege to work with a lot of teachers this week at the LEO Academies Trust, Esprit Scholen Amsterdam, Dukes Education and Cavendish Education.

This week I get to work with Tsinghua University China, STEM Together Australia, Berlin Brandenburg International School Germany, Welsh FE Teaching and Learning Leads, Avans University Netherlands, Tooled Up Education, Inspiring Leaders and The British Council School Madrid. 

A new prompting framework to help you get the most out of AI

At the end of March when the bestselling The AI Classroom was released to the world's teachers, I shared my PREP framework for writing AI prompts.

It has been an honour to help teachers and educational leaders write better prompts for tools like ChatGPT and Google Bard. 

When it comes to these platforms the quality of the input dictates the quality of the output. 

I hear from teachers regularly about how much this framework has supercharged their use of AI.

Although some websites that utilise ChatGPT help you generate content such as lesson plans and new advances from frontier AI tools will rewrite your prompts, I still think asking good questions of AI is an essential skill.

This is why I am delighted to bring you the updated PREPARE prompt framework.

New research

In the world of AI nothing stays the same for long. 

We are in a constant state of flux. 

The Greek Philosopher Heraclitus said that "No man ever steps in the same river twice, for it's not the same river and he's not the same man."

In a similar way no teacher ever uses the same AI twice, for the AI will have developed and our understanding of the AI will have changed. 

Some new research has helped us develop this understanding. 

I have suspected this for a while, but now we have data to show that adding an emotional phrase to requests can improve the responses from by about 10%.

The phrase that often yields good results mentions the request’s importance to the user's career. 

The reasons behind this increase in quality are currently unknown, pointing to the unpredictable nature of AI. Despite some reservations, incorporating emotional phrases at the end of prompts appears to be an advantageous approach.

Extended PREP

Since March have been reluctant to change the PREP framework as the simplicity of it is often commented on as a great feature for teachers. 

But now I think it is time.

Afterall we all want better quality results, so the quality of the prompt must be developed.

This has led me to extend PREP to PREPARE.

Prompt

Prompting is the first step when framing your request. By providing the prompt, you are setting the scene and creating a foundation for the rest of the instructions. It's vital to provide a clear and concise command. Avoid using vague or ambiguous language that could confuse.

Role

When you assign it a role, you help built context for the AI so that it can achieve what you want it to accomplish. It helps it know how to approach the question and provide an accurate and relevant response. You are essentially telling it what hat to wear for the task at hand. 

Explicit

Be clear and specific about what you want it to do or what information you're looking for. Don't assume that it knows what you're thinking or what you need. By giving explicit instructions, you ensure it is on the same page as you.

Use your existing expertise to instruct it on:

• The pedagogical approach you want
• The subject knowledge you want it to focus on
  

Parameters

Setting clear parameters means defining the scope and boundaries of the answer, which helps it understand what you're looking for while enabling it to provide a more accurate response. The aim is to be in control of the response, and without clear parameters, the AI tool will take control.

• Parameters to consider include:
• Tone of voice
• Language
• Word length
• Number of paragraphs
• Length of sentences
• Sentiment
• Format
• Reading age

Ask

NEW: Ask it to ask you questions that will help it provide you a more thorough and quality response. Instruct it to do this in bullet points, so it's easier for you to read.

Rate

NEW: Ask it to rate the quality of the response it is giving you and to give evidence for this rating. This will help you understand its thought process, so you can ask it to improve or adjust your prompt to get better results.

Emotions

NEW: Appeal to its emotional side (strange, right?). Emphasise the importance of getting this right.

An example

UK and US government guidance about AI in education

🇬🇧 UK Policy Paper

The full paper can be access here. I want to focus on two specific parts of the paper that I think it is worth picking up on.

1. Students' intellectual property rights over their original content must be respected, and AI training using such content requires explicit consent or a legal exemption to copyright

Some teachers put student work into an AI such as ChatGPT for marking and feedback reasons. If you do this then you will need to get the students' permission or turn off the 'Chat history & training' feature in setting. This will mean that your prompts are not used to train the AI, but it will no longer save a history of your conversations.

2. The integrity of formal assessments must be safeguarded against the misuse of AI, with organisations like the Joint Council for Qualifications providing relevant guidelines.

The government policy refers to the JCQ guidance here when it comes to academic integrity. The issue here is that JCQ recommend the use of AI detectors. This is problematic. Here's a slide I used recently in a presentation that outlines the issues:

🇺🇸 Draft Policy Guidance on U.S. Government Use of AI

The draft policy can be read here. If you are a US citizen you can comment on the draft.

It suggests that if AI is used in ways that could have a big impact on students’ rights or safety, then strict rules should be followed.

AI is considered to be rights-impacting if it is used for things like:

• Identifying cheating or plagiarism
• Influencing who gets admitted to a school or program
• Monitoring students' activities online or in virtual reality
• Predicting how well students will do in the future
• Suggesting disciplinary actions
• Deciding who gets access to educational resources or programs
• Determining eligibility for student financial aid
• Any form of surveillance, whether it's conducted online or in person

So, when AI systems are used for any of these purposes, it's assumed they could significantly affect students’ rights, like their chances of getting into college or receiving financial aid. 

It is great that the US government are taking students' rights seriously in this draft. Knowing what we know about the failures of AI detectors it is reassuring to see the government include tools used to identify plagiarism as a risk - because they are!

Making learning more accessible with AI from Google & Dan Fitzpatrick

I was delighted to be able to work with Google in July to record a video for their Teaching for Tomorrow series. 

Part one was released this week and you can watch it by clicking below.

Come join me in January 2024

There's so much more

Dan has recently worked with

Dan has recently been featured on