LISTEN BACK TO THE FULL MASTERCLASS

We’ve created a podcast version of the session. It captures key points from the session including the generative AI landscape, real-world examples, platform comparisons, and our RISEN prompt framework – all in audio form.

Created using ChatGPT and NotebookLM, click to listen to the podcast:

WHAT IS GENERATIVE AI AND HOW DOES IT WORK?​

Cutting through the jargon. Here’s what GenAI really means

Generative AI is a specific branch of deep learning focused on creating new content – text, images, audio, video – based on patterns it’s learned from huge datasets.

This is made possible through models like:

• LLMs (Large Language Models) including ChatGPT or Claude, which generate human-like text
• Diffusion Models for high-quality visuals and video (e.g. Midjourney, Sora)
• Neural Networks for realistic audio and voice (e.g. ElevenLabs)

THE OPPORTUNITY IS HUGE, BUT SO ARE THE RISKS

As GenAI adoption accelerates, so too do concerns around data, IP and brand safety.

• Capgemini found that 97% of companies using GenAI have reported a related breach or incident
• One platform (OmniGPT) is reported to have leaked 30,000 users’ information
• Research from Deloitte found that deepfake scams contributed to $12bn in fraud in 2024 alone
• Around 40% of users have shared confidential info unknowingly, according to Indusface

Add to that the murky waters of copyright (most models are trained on copyrighted content), and you’ve got a clear mandate: businesses need to know the risks before jumping in.

This isn’t just about productivity, it’s about protection.

LISTEN TO THE HIT TRACK WE CREATED USING AI

“Five Hundred in Five (Dunford’s Drive)” was written, produced and performed using a combination of Suno AI and a Custom GPT. The entire process, from brief to beats, took minutes.

Think of it as proof of concept: GenAI isn’t just for words and images. It’s a creative engine that spans sound, storytelling, and emotion.

FOCUS:​ CHATGPT​

Multimodal creativity, planning and problem-solving

ChatGPT is still the most widely used generative AI tool for good reason.

Whether you’re writing copy, planning campaigns, brainstorming names or building workflows, it’s fast, intuitive and incredibly versatile.

The most widely used version, GPT-4o, supports not just text but also voice, images, vision and code, all from a single interface. It also connects with tools like browsing, file uploads, image generation and more.

It’s not just a chatbot. It’s a creative and strategic co-pilot.

Custom GPTs: ​

Tailored AI assistants, built around your own content

A Custom GPT is a personalised version of ChatGPT – one that behaves the way you want, speaks your language, and draws from your own documents or data.

You can:

• Set clear instructions for tone, behaviour and focus
• Upload PDFs, presentations or manuals it can reference
• Enable tools like image generation or browsing
• Share your GPT with others – or keep it private

There’s no coding required. You build it through a simple interface and it’s ready to use instantly.

FOCUS:​ SORA​

AI video generation – animate your ideas in seconds

SORA is OpenAI’s powerful video generation platform that can transform images or prompts into short, high-quality clips – no cameras or crew required.

You can generate up to 20 seconds of video at 1080p resolution, either by:

• Uploading an image to bring it to life with movement, or
• Providing a text prompt to create something from scratch

It’s ideal for prototyping creative concepts, visual storytelling, or early-stage campaign testing. Whether you’re mocking up an ad, animating an idea, or building out a brand world, SORA makes it fast and surprisingly cinematic.

FOCUS:​ Perplexity​

AI-powered search with real-time, cited answers

Perplexity is an AI-driven search engine that merges the immediacy of traditional search with the depth of large language models. Unlike conventional search engines that provide a list of links, Perplexity delivers direct, concise answers sourced from multiple reputable websites.

It’s like having a research assistant who can:

• Pull live results from across the web
• Reference academic papers, YouTube transcripts, and more
• Show sources for every answer it provides
• Track your follow-up questions in the same thread

Great for competitor insight, market trends, regional behaviours or getting up to speed on a new topic quickly, and with confidence in the source.

Focus: NotebookLM​

Turn your materials into structured thinking

Where ChatGPT is brilliant at idea generation, NotebookLM is built for making sense of complexity. It lets you upload your own content – from PDFs and websites to video links and notes – and then works across those sources to help you understand, summarise and connect ideas.

It can generate:

• Mind maps and structured outlines
• Briefing packs and summaries
• Quizzes, podcasts and more

It’s especially powerful for repurposing content, building internal knowledge hubs, or preparing fast, smart outputs from messy input. If you’ve got information overload, this is the tool that brings it all together.

THE POWER OF THE PROMPT

The prompt is your map and AI only follows directions

Generative AI tools are powerful, but they don’t read your mind. The real value lies in how you guide them. That’s where prompting comes in – the way you structure your request has a direct impact on the quality and relevance of the output.

Most users still fire off vague prompts like:
“Write some ad copy for a new sports shoe.”
And yes, you’ll get something… but not necessarily what you wanted.

That’s where RISEN comes in.

A simple, practical framework designed to structure your prompt for better results: