The Plain English Guide to: Generative AI  

Introduction 

Generative AI is an overnight success story decades in the making.  

It may seem as though it came out of nowhere in November 2022 with the wildly successful launch of ChatGPT, but generative AI (genAI) has been in the works for a lot longer than you might think. 

There’s also some confusion, especially in the business world, around how genAI can help us do our work better-as opposed to it “stealing” our jobs-and what the limits of its abilities might be.  

So, to clear things up a bit, let’s take a closer look at generative AI and explain exactly what it is the Get Support way… in plain English.  

What is generative artificial intelligence? 

Generative artificial intelligence refers to technology which can produce new content-text, images, and even video-by learning patterns from existing data.  

The tech is built on Large Language Models, or LLMs, the most common today being Generative Pre-trained Transformers, or GPTs. Unlike traditional AI, which generally analyses data to find patterns and make decisions, generative AI takes things a step further by actually creating something new. 

Importantly, it can’t think for itself or generate anything outside of what it’s already learned, meaning human creativity is still essential for producing the best results from genAI tools.  

Of course, with great power comes great responsibility.  

Generative AI can sometimes produce results that are unexpected or even biased, as it mirrors the data it’s trained on. That’s why it’s crucial for businesses to pair these powerful tools with human oversight, ensuring the outputs are not only innovative but also aligned with your values and standards. 

Where it all began: A (brief) history of generative AI 

Even though it feels like genAI is the new kid on the block, it’s been a long road to get to the advanced genAI models we use today.  

Here’s a quick timeline of the key milestones in the history of genAI development: 

• 1950s. The inception of AI as a concept, with Alan Turing’s pioneering work on machine intelligence. 
• 1980s-1990s. Development of neural networks, which laid the groundwork for modern AI. 
• 2000s. Advances in machine learning algorithms and the emergence of deep learning. 
• 2017. Introduction of the Transformer model by Google, revolutionising natural language processing (NLP). The Transformer model led to the Generative Pre-trained Transformer, which led to-you guessed it-ChatGPT.  
• 2022. November 30^th of 2022 saw the launch of ChatGPT by OpenAI. In terms of generative AI as we know it today, this was a red-letter day and marks the beginning of the AI boom.  
• 2023. Microsoft introduced the Copilot product line, including Microsoft 365 Copilot, which integrates generative AI features into the core Office apps like Word and Excel. 

How does generative AI work? 

It’s easy to get confused by the amount of tech jargon which goes along with genAI. LLMs, GPTs, NLP… the list goes on. 

So let’s break down generative AI by way of analogy.  

Consider a master chef in a world-class restaurant. Over many decades, this chef has learned from thousands of recipes and is now able to use that experience to whip up new, delicious dishes from scratch. They won’t be identical to the recipes the chef has made before, but they will share characteristics with them.  

In a similar fashion, genAI has learned from vast amounts of data-essentially the entire internet-to understand patterns, structures, and context. It then uses this knowledge to create new content which makes sense and fits the context. 

The LLMs at the heart of genAI are trained on diverse datasets, absorbing grammar, facts, and visual data. When user writes a prompt for a genAI tool, the model uses its training to generate relevant and coherent responses.  

Best of all, because LLMs are built on natural language, working with the tool is as simple as having a conversation. That means practically anyone can work with a genAI tool without the need for complex coding or previous experience.  

The most common generative AI tools for small businesses 

You’ve probably already been bombarded with new AI tools and AI updates for existing business apps, so let’s cut the wheat from the chaff. 

Here are some of the most popular genAI tools for small businesses: 

• ChatGPT. Developed by OpenAI, ChatGPT is known by everyone and their grandma at this point. Built on OpenAI’s GPT 4 model, it’s a versatile chatbot which can assist with everything from holiday ideas to business strategy.  
• Claude. Created by Anthropic, Claude excels in providing human-like interactions. It’s particularly useful for detailed and nuanced conversations, making it a great tool for support roles. 
• Google Gemini. Beginning life as Google Bard, this tool integrates Google’s advanced AI capabilities into various applications, offering powerful solutions for text generation, summarisation, and more.  
• Perplexity. Combining the expansive knowledge of an LLM with the accuracy and recency of a search engine, Perplexity.ai is a handy tool for research (or just going down internet rabbit holes).  
• Microsoft Copilot. Integrated directly into Microsoft 365 apps like Word and Excel, Copilot enhances productivity by automating repetitive tasks, generating content, and offering insights. Because it’s built into existing apps rather than separate software, it makes genAI accessible anytime.  

Will generative AI really steal everybody’s jobs? 

Now for the big question: is generative AI really here to take our jobs? Can AI ever be a replacement for human effort? 

The way we see it at Get Support, the idea that AI will steal jobs is hugely overblown.  

While it’s true that AI can automate certain tasks, it’s far more likely that genAI will act as an accelerator of effort rather than a replacement for it. Think of it as a new type of tool-just like the internet or smartphones-that will be integrated into the apps and systems we already use. As discussed above, we’re already seeing genAI integrated into the apps we already use every day, and that’s likely to continue. 

But if you’ve spent time working with genAI tools, you’ll already know that the output is always missing something-and that’s the human touch.  

If you’d like to learn more about Microsoft 365 Copilot, or give it a try in your organisation, just ask your Get Support Customer Success Manager or call us on 01865 594000.