Oops! We could not locate your form.
AI-generated content is in every marketer’s mouth these days. But what does it actually mean? What can this technology do, and where might its limitations lie? Join us for an unbiased exploration of one of the most controversial topics in marketing today.
As a marketer, it’s almost impossible to avoid. Everywhere you look, you’ll find someone talking about AI-generated content. Podcasts, industry blogs, social media, and conferences have made it a banner topic, and for a good reason.
For some, it’s the greatest thing since sliced bread. For others, it’s a first look into the coming marketing (and ethics) abyss.
Which of those two sides is right? That might be impossible to answer for now.
One thing we know for sure is that it’s a controversial topic, and it’s not going away anytime soon. Just look at the latest buzz surrounding ChatGPT, which crossed 1 million users within a week of its launch date and can apparently pass law school exams.
Love it or hate it, AI-generated content and tools like ChatGPT are here to stay. So let’s take an unbiased look at the technology, from its basic capabilities and limitations to the current controversies and what the future might hold.
At its core, the content generated through artificial intelligence is simple. A user inputs a question or phrase, similar to how they would in a search engine. But instead of simply finding existing results, the tools piece together bits of information from the data sources they’ve been for an ostensibly original answer.
So far, we’ve seen the buzz around AI-generated content in two basic applications:
The current promise of both of these types of tools is self-evident. A simple query turns into comprehensive content after just a few seconds wait. Theoretically, entire marketing tasks like content marketing can be all but automated, with the writing done by the computer rather than a human.
These tools have the potential to generate content at scale. No more waiting for the right writer to write the right post or an artist that needs the inspiration to strike. It’s all right there, at your fingertips, ready to go within seconds.
Savvy marketers already know that the current promise of AI-generated content sounds just a little too good. Even the name is a tad misleading. It isn’t truly AI (artificial intelligence) at work, but instead machine learning algorithms, and things like natural language processing. It’s all very clever but not quite Turing Test ready yet.
The actual tools of today come with some pretty significant limitations that have caused an entire industry to tiptoe around them. Even ChatGPT which is seeing immense popularity, has some fairly stark limitations.
Most obvious are these 5 that every marketer even considering the use of these tools should keep front of mind:
No content created by a machine is actually original. It’s all generated by pulling together elements from different sources the machine learning algorithm has been fed and outputting them into one coherent piece of content.
That output might be impressive, but it’s still a nice-looking Frankenstein’s monster. Here lies the problem. AI tools don’t cite their sources in their output, so that lack of transparency and originality can directly lead to plagiarism. Particularly where those whose content is drawn on by the tool to create its own content, have not consented to have their work used in this way.
If you fed a machine learning tool images of every painting in the National Gallery and asked it to create an Impressionist-style portrait of a man staring pensively out of a window, you’d definitely recognise stylistic and compositional choices from the world-renowned artists it was given to learn from.
Now give it the whole of the web, and it has access to the works of every creator who has work shared online. From someone who sells prints on Etsy, to Degas.
Some argue that this isn’t plagiarism but rather, inspiration, and that all creators are influenced by other content they’ve consumed. Others argue vehemently that it is plagiarism and none consensual use of their work.
This goes for written content too. If you created a unique and interesting study in a field you are an expert in, you’d expect to see it referenced and credited in the content generated off the back of it.
From a purely practical standpoint, AI-generated content lives and dies on the data it is provided with. For tools like ChatGPT, this is a major stumbling block.
Whilst people out there are singing its praises as if it is the second coming, and making bold claims about its abilities, the fact is that ChatGPT doesn’t have access to the live web. It is using a snapshot of the internet from 2021.
So using it to create factual content on any subject that experiences frequent change is problematic. No doubt you’ve seen some examples of it giving very plausible but very incorrect responses.
AI gets things wrong. Quite frequently, in fact. It can’t say what the first TV cartoon was, or who created it. Forget about correctly attributing the founding of the first iPod.
As of right now, there is no real fact-checking mechanism built into the most common AI tools, just user feedback (often abused) meaning the content you are provided, often designed to build your credibility, is not trustworthy.
Because AI-generated content is essentially crowd-sourced and regurgitated, it’s impossible to use AI tools for truly original content. That, in turn, makes it impossible to develop a content strategy to set yourself apart from your competition. You won’t be able to get your unique brand voice or any unique content that makes you stand out. Nor that unique understanding you have of your audience. Your content will be akin to any other AI-generated content that draws on the same data and language models.
What’s more, as the web becomes more and more heavily populated with AI-written content, these tools will be generating content from content that is already AI-generated.
There is no shortage of controversy surrounding AI content at the moment.
Google recently posted an article explaining how it treats automated and AI-generated content. Where previously it stated that it will devalue AI content it can detect in its search rankings, after the announcement of BARD, they have now clarified their stance.
It is no surprise that between the recent Helpful Content and Spam updates, many sites heavily utilising AI-generated content experienced big rankings hits.
Google want to present searchers with the best factual content, written by people with expertise and experience. They also have their own AI and natural language processing models to help identify this type of content.
Always evaluate your content and reasons for producing it (whether you have used AI or not) to ensure it aligns with Google’s definition of high quality content if you hope to have it rank.
However content is produced, those seeking success in Google Search should be looking to produce original, high quality, people first content, demonstrating qualities E.E.A.T.
Google – 2023
Prestigious publications are seeing PR hits when using AI art for their covers and articles. We’re even seeing the first lawsuits brought against some of the world’s biggest organisations for plagiarism through AI content generation.
In the last couple of days, a group of female Streamers discovered that their likenesses had been used to create AI-generated pornography, without consent, and to generate money. The site has since been taken down but is a clear example of the huge issues that lay ahead with this technology and the lack of regulation currently in place.
In other words, as AI-generated content is becoming more impressive, it is also becoming more controversial by the day. That’s before we even begin to discuss some of the wider ethical considerations floating around it.
To say that AI art has been controversial among artists might be an understatement. Because software like Lensa pulls information from other sources, artists have voiced strong privacy and copyright concerns:
Artist Mateusz Urbanowicz, whose work was also included in LAION-5B, said that fans have sent him AI-generated images that bear striking similarities to his watercolour illustrations. It’s clear that LAION is “not just a research project that someone put on the internet for everyone to enjoy,” he said, now that companies like Prisma Labs are using it for commercial products.
“And now we are facing the same problem the music industry faced with websites like Napster, which was maybe made with good intentions or without thinking about the moral implications.”
The legal landscape may be murky, but the strong voice of artists is beginning to have reverberations. Getty Images, for its part, has banned all AI-generated imagery. It’s a legal minefield that has caused many experts to recommend against brands using the practice altogether.
Combine the potential uses of AI content with its considerable drawbacks, and one conclusion is clear. This is a battlefield that won’t soon be decided one way or the other. A look to the future clarifies some things, though uncertainties still remain.
Some experts predict that nearly all web content will be AI-generated by 2025. Others point to a more optimistic future in which, used correctly, the same tools will actually drive creativity.
No one knows exactly which of these predictions will come true, or if either of them will. But of course, that uncertainty alone makes this topic worth monitoring for any marketer.
In the meantime, there are some responsible uses of AI content that should not be ignored.
While pure content creation may be off-limits, it can still be used:
One thing is clear. The future of marketing will include AI-generated content in some shape or another. The key, then, will be not avoiding it altogether, but finding ethical ways to leverage it for enhanced productivity and better promotional outcomes.
It’s impossible to outline all of the nuances, ethical implications and controversies of AI-generated content in a single post. Its potential opportunities and considerable issues are just too complex.
Still, it’s a topic that will only continue to develop in future months and years, and every marketer should keep an eye on it for potential implications.
If you’re looking for more insights into marketing, web development, and other digital topics, explore our blog.
Discover the power of Enhanced Conversions in Google Ads. Learn how this feature can help you optimise campaigns while respecting user privacy.
For the majority of businesses, having a strong online presence is crucial.
Come July 1st 2024, access to Google Universal Analytics properties and the API will cease. With it will go any historical analytics data contained within your properties.
