AI-Generated Content: Should You Ever Use it?

AI is a hot topic right now. Doom-mongers and utopians have been talking about its advent for years, but things took a giant leap forward for robotkind with the release of ChatGPT.

This bot-driven content generation tool has been wowing users with its abilities in spheres hitherto felt to be areas of human dominance. If content is king, it seems suddenly like we have a new kingmaker. One example of AI’s capabilities is using GPT-3 for SEO to optimize content for search engines.

So, now the question has changed from ‘what can AI do?’ to ‘what do we want it to do?’. Are there fields of endeavor that we want to remain purely human? Or is there an advantage to be had in securing the assistance of AI to produce quality and original content?

What is AI-Generated Content?

AI is used in a colossal range of applications now, from the Dialpad cloud PBX system to weather forecasts. In the field of content generation, it’s making some especially exciting inroads.

When material is produced not by a human writer but by a large language model (LLM) with advanced creation properties, this is AI-generated content. Artificial intelligence needs a prompt, in the shape of a question, instructions, or series of keywords, and a piece of content will result.

Until fairly recently, language models tended to be limited to short-form answers to specific questions. With larger, more sophisticated language models such as those used with ChatGPT, there’s been a monumental widening of what AI-generated content can mean.

What was often all too obviously artificial in origin can now quite readily pass for human-written content. This means that Alan Turing’s famous test might at long last have been passed.

But before we break out the bunting and festoon our computers with ticker tape, we need to pose questions about the desirability of what’s been achieved.

What are the Advantages of AI-Generated Content?

1. It can kick-start the content creation process

The fact is, human content creators aren’t machines. This one’s anything but. Of course, this may be an advantage. But it can also seem like a huge hindrance when a writer’s sitting there staring at the cursor, waiting for inspiration to strike.

It’s so much better for the writer, the editor, and the reader if the human writing ignition key can be turned on and things can then proceed with flow and fluency. AI can be good for this. Give it a few keywords and see where it goes with them.

It might give you ideas that you’d never have thought of. You can then develop them in a very human tone of voice, knowing what your readership is and how best to communicate with them.

2. It can save money

When you have a multitude of platforms for which to produce content, it can mean you have a whole army of writers who need to be paid. There are precious few writers who will do the job just out of a love for the written word.

With using AI tools, you can take a piece of content and automatically adjust it to fit the parameters you need to hit to perform on numerous different platforms. In this way, the power of automated text ads, for instance, can be harnessed alongside a myriad of other content marketing efforts.

3. It’s good at dull

Sometimes, a marketing writer churns out product copy that merely describes an item without having to be exciting or persuasive. This is the kind of thing that AI excels at: informative, uncontroversial writing, like an AI-generated Dialpad small business VoiP system features rundown.

This then frees human writers to concentrate more on communicating passion to their fellow humans.

What Are the Disadvantages of AI-Generated Content?

1. It’s not original

For all the plaudits showered on AI-generated content, it’s derived purely from what’s already out there. It’s pretty much a collection of what the bot’s managed to find on the net, assembled in what it perceives to be a satisfactory format.

This means that it’s good for coming up with helpful content like answers to straightforward questions and routine blog posts. OK, the questions that it can deal with are getting more and more complex as time goes by (even tax queries, albeit not too tricky ones), but AI will struggle to produce high-quality content such as convincing long-form opinion pieces, for instance.

2. You don’t own the content

The latest guidance on AI content ownership was just released on March 16th, 2023 from the United States Copyright Office. They were transparent about the fact that there is still work to do on their end to formalize their policies, but their current position is as follows:

In the Office’s view, it is well-established that copyright can protect only material that is the product of human creativity. Most fundamentally, the term “author,” which is used in both the Constitution and the Copyright Act, excludes non-humans.
– United States Copyright Office

In summary, regardless of the “human creativity” that you many have used in developing a prompt that generated unique content, the content output was generated by the AI tool. Therefore, it cannot be copyrighted. Now, does that mean you can’t use it? Well, that really comes down to your situation.

If you are a content producer for a company and that company expects that all published content is protected by copyright, then you shouldn’t be using an AI tool for content generation. If you work for an agency and your client contracts dictate that anything you produce for a client is owned by that client, then you shouldn’t be using an AI tool for content generation.

3. It’s not always correct

As mentioned, AI gets its information from the internet, and — spoiler alert — not everything on the internet is true. While AI can attempt to seek authenticity by trying to back up its findings with multiple sources, its tendency to get things wrong is worrying. It’s concerning for the miscreant schoolchild using it to produce an essay, and downright alarming if AI gets used much in strategic defense early warning systems.

In time, there may be a system of authentication put in place so that more credibility can be secured. Just like a data engineer being able to brandish Databricks certification AI might achieve trust this way.

Something that’s of particular interest here is an unforeseen AI characteristic: ChatGPT makes stuff up. It’s not that it’s giving you bad information that it thinks is correct. It can sometimes invent information based on what it predicts you are looking for. This is a product of a programmed tendency to seek approval.

While on the one hand, this is almost endearing and a little human-like, it’s not much use if the user thinks they’re getting 100% rock-solid accurate data.

4. Google can object

Google's policy on ai-generated content

There have been numerous instances of search engines such as Google taking exception to using black hat tactics to improve search rankings. Unfortunately, it’s this kind of material that AI tends to major in. So, if you’re using AI to generate material, it’s well worth casting an eye over it to make sure it’ll pass muster.

So, When Should You Use AI-Generated Content?

AI has very specific strengths but also inherent weaknesses that it can bring to the writing process. What this means for a content strategy is that AI should be used in tandem with human writers. That way, mistakes can be reduced and Google is kept happy. It will also result in happier readers. This is because, despite these revolutionary improvements, AI-generated text can still sometimes come across as a little stilted and downright odd.

How things will pan out in the future is a little unpredictable. AI will certainly become more and more creative in due course, and AI-generated writing will become increasingly commonplace. Whether this means human writers will ever be phased out, only time will tell.

For more great tips, tricks, and how-to’s, subscribe to the DivvyHQ blog today!