If recent times have taught us anything, it’s that artificial intelligence (AI) will soon be playing a major role in our lives. Whether people are creating art via programs such as DALL.E 2 or generating copy through ChatGPT, there’s a reason that AI is dominating the headlines.
But is it time to surrender the creative process over to our robot overlords? And what does AI mean for the future of communications?
The creative applications of AI are undeniable. Built and trained using extensive data sets, these programs instantly draw on information from millions of sources to generate material that seamlessly imitates that created by humans. The biggest benefit of this is, of course, speed. AI is already being used by companies across the globe to create content to free up time for other activities. Want an article written about pension policies? An AI program could create one in mere minutes – and it probably won’t be the worst thing you’ve ever read, either. You can also use certain AI tools to tailor existing content for different audiences and markets with ease.
AI is also transforming visual assets. Need a specific image for a website but can’t find one? You could generate a believable picture of anything you can imagine whether that’s a cat building a sandcastle or a watermelon performing an opera. Effectively, there are no limits to what AI can create, which makes it a very appealing prospect for those who need unique material.
AI can also help companies save on costs. Why hire a talented artist, writer or animator if you can have AI generate similar content in a fraction of the time? In reality, we could have used AI to write this whole article and populate it with appropriate images - but there’s a reason we didn’t. A few reasons, in fact.
Our experiences are what make us human. A creative team isn’t just valuable because of a set of skills – we all bring our own perspectives, ideas and innovation with us, meaning that we can create content that exceeds expectations and that truly makes an impact.
We understand the human side of every project and can feed this into what we create to ensure that every message resonates with the intended audience. As sophisticated as AI is, you can’t form a working relationship with it, and there’s no guarantee that what's generated by AI will fully follow your vision.
Humans aren’t perfect and often make mistakes. However, when this happens, they’re usually simple errors that need minor adjustments. Correcting AI generated content is often just as time-consuming as creating our own content from scratch.
This is especially true for images – an AI generated picture of people in an office may be fine at first glance, but look closely and you’ll often find that people have extra teeth or strange hands. Perfect if you want to lend a certain Lovecraftian je nes sais quoi to your campaign, but often not the impact you’re looking for.
AI copy also has imperfections. It’s often repetitive and continually follows the same structure, leeching out all personality and creating communications that sound like they’ve been written by, well, a robot.
As AI programs gather data from so many sources, there is significant concern that using these programs to generate content could lead to copyright issues further down the line. In some instances, these programs are plagiarising existing work to create new content – something we don’t tolerate from humans, so why would we tolerate it from computers?
There has already been huge backlash from artists whose style has been clearly ripped off by AI programs, and this is an issue that is only likely to become more widespread in the future, especially if no adequate protection is offered to artists. At the time of writing, there is no real legal clarity surrounding AI content and who owns it, which means treading on uncertain ground if you choose to use AI for content generation.
It’s easy to forget that AI has an environmental impact. However, training these models is an energy-intensive process. Research suggests that the average carbon footprint generated by training an AI model is as much as 284 tonnes of carbon dioxide, five times the emissions generated during the lifetime of a car.
That’s not to mention the materials used to craft these AI systems and the ongoing energy usage necessary to maintain and dispose of them – there’s a reason that current environmental conversations are implicating AI as furthering climate change.
As with all emerging technologies, it’s impossible to know the full extent of how game-changing AI will be. The benefits these models present are undeniable, especially if they continue to become more sophisticated and able to iron out several of the perceived negatives.
However, what’s clear is that people are still at the heart of what we do as communicators, and AI won’t change that. There’s plenty of scope for these tools to be used as part of the creative process, providing they are used thoughtfully and in the right way by creatives who are masters of their craft.
i New Scientist. Creating an AI can be five times worse for the planet than a car. Link: Creating an AI can be five times worse for the planet than a car | New Scientist [Last accessed April 2023].
ii Dhar, P. The Carbon Impact of Artificial Intelligence. Nature Machine Intelligence volume 2, pages423–425 (2020)
