What is Generative AI?

Everywhere you look, news about generative AI platforms like Chat GPT abound. But the technology isn’t new, so much as it’s gotten amazing press in recent months.

We created the cover image of the article with the help of DALL-E. The input command (Prompt) was as follows: “Please generate an image in 16:9 format. It should show you in an environment in which you feel comfortable.”

So, if you’re left wondering what generative AI really is and why it’s so popular now; we’ll tell you all about it in this quick read.

What is AI?

First, it’s helpful to know what it is generally. Well, it stands for artificial intelligence and it’s a term we humans use to describe synthetic smarts i.e. machine learning that can do tasks – with or without the help of people.

And you probably use it every day, if you have a virtual assistant like Siri on your phone or Alexa in your home.

What is generative AI?

Generative AI is a flavour or type of artificial intelligence that can make things. You’ll see it in apps or software that create songs, text, images, videos and more. It’s the next layer that goes beyond just making sense of data and adds on creation. But it doesn’t have experiences of its own (yet). So, by definition, everything an AI can create is referential to human artistry.

It can’t really understand you and everything it knows is based on things it was taught by its creators or patterns it found within data supplied by people.

How does generative AI work?

The way generative AI works is truly remarkable, however. Even if it isn’t ‘true creativity’ like we’d ascribe to Picasso or Monet, the things it can make in record time are pretty extraordinary.

First, like most AI, it’s trained on a large dataset. This could be selected images or text up to, and including, the whole of the internet. It mines that data to find patterns or clusters and then seeks to make ‘new’ content based on its training data.

What is Generative AI vs Machine Learning?

AI is the overarching category in which machine learning and generative AI sit. Generative AI uses a type of machine learning called neural networks to make sense of all the data it has access to. The generative part is what it does with it – creating new text, audio, video, image or data content. And these systems can already make some pretty impressive things.

What can generative AI do?

Every week, it seems, we’re learning about new things that it can do. It can write code for a new video game, compose a song, write a book and create a digital painting in seconds.

In business, it can analyse vast amounts of data to make customer support more intuitive or suggest the next product innovation based on market data and consumer buying patterns.
It can help workers respond to emails faster or draft formal letters of complaint to your local council. Currently, we’re still testing the outer limits of this technology and we’ve not reached the ceiling yet.
At Qymatix, we use machine learning for AI-based sales forecasts with predictive sales analytics, although this does not fall within the scope of generative AI.

What are some examples of generative AI?

Well, you’ll already know about Chat GPT – the search bar meets virtual assistant that can speed read, create content and suggest ideas faster than any human ever could.

But there are so many other tools out there that use generative AI. First, there’s Bard, Google’s version, that’s integrated with their suite of cloud tools.

Then there’s DALL-E, an AI image generator from the same folks behind Chat GPT that got big press a few years ago. We also have Grammarly, an AI-supported and pratical writing assistant.

In fact, new tools are announced all the time – rising to prominence as people use them to make headshots for work or plan their next holiday.

Why is generative AI so popular now?

Generative AI has just been given the spotlight due to the virality of Chat GPT, but it has been used in a smaller capacity for many years now in areas like spam detection and basic copywriting.

As companies begin to invest in and expand on the technology (like with Microsoft’s new Copilot) the opportunities to use generative AI within your own business and personal life will grow as well.

What’s the economic impact of generative AI?

The economic impact of generative AI has not yet been realised fully. Demand for this technology and the people who know how to use it like Prompt Engineers has seen new sectors open up. However, generative AI does mean that some professions are at risk like those in media, tech, market research, legal, teaching, finance and design.

Anything that’s formulaic with predictable outcomes and a lot of prior precedents is fairly easy to translate over to non-human labour – provided there are no downsides to doing so.

But, before you run off and use a generative AI tool to write copy for your whole company website, you might want to be aware of the risks and limitations.

Are there any risks with generative AI?

Right now, the biggest risk to using open-source generative AI is a lack of transparency and citation.

When you type a prompt into Chat GPT and get a response, you don’t have any information about where its conclusions came from. You don’t know if it has any biases or even if the responses it gives are correct.

Plus, since computers can’t have own experiences, anything that a generative AI creates will be based on real human input. And that has a lot of people worried about copyrights and plagiarism.

It can also be harnessed to do harm like creating spam, delivering propaganda, making malware or crafting deep fakes that damage reputations.


What is Generative AI? – Summary

Although there are risks associated with generative AI, these depend very much – as with any technology – on the implementation and the user’s intentions.

Artificial intelligence – whether “normal” AI or generative AI – is a technology that offers enormous opportunities and is constantly growing. Now is the time for companies to use AI wisely and stay one step ahead of the competition.

If you’re curious about how AI and machine learning can help you gain an advantage and make more data-driven decisions faster than ever before, let’s talk.