It’s hard to ignore AI (Artificial Intelligence) at this point. Either you are using it, talking about, protesting it, or at least seeing videos on it constantly. The AI bubble is here and shows no sign of popping anytime soon. The topic can feel overwhelming at times with the amount of information available at your fingertips. So, let’s focus on just one aspect of it today and talk about the rise and impact of AI on content creation.

The Rise of AI in Content Creation

AI and content creation have been working together for much longer than many would like to admit. AI programs have been keeping content creators’ grammar in check long before the ability to generate a video of a talking dog on a surfboard. Recently, though, the rate at which new features and programs as well as what they are capable of doing has been growing at an exponential rate. It seems every month, companies like Adobe, Open AI, Canva, and others are rolling out new features and integrations. The big call out, though, is that most people assume AI means Generative AI. However, at the end of the day, AI just refers to a program that has been trained to mimic or replicate human intelligence. Generative AI creates new content like text, images, or code from patterns it learns, acting as a creator. Not everything a content creator uses is Generative AI vs traditional, so it’s important we look at the whole picture when having this conversation.

How is AI Being Used for Content?

Most AI used in the content creation process can be broken down into two categories: “Polishing” or “Producing”.

Polishing: When the AI program is being used to enhance an asset or simply assist the user by simplifying a tedious part of the creation. Examples of this would be grammar checking services, background detection, auto adjustments of photo touch ups, enhancement of audio, etc.

Producing: When the AI program is creating an asset from a prompt. Examples of this would be generated images, videos, music tracks, and copy.

While considering the above definitions, it may be easy to say, “So Producing is just Generative AI and Polishing is everything else.” There can be a lot of grey area and crossover within the two categories. For example, if you typed an entire script and used a service that generated a voiceover of the copy that you wrote, were you polishing or producing? You can argue both, and the two aren’t mutually exclusive. While you did generate something, it wasn’t as in depth as telling software to create a speech on a topic. In this case, I’d lean this towards this being slightly more polished. It’s an additive element to something that you are still curating and editing. It is not replacing the process of making the whole asset but rather assisting with a part of the process. This nuance is important to consider when deciding whether or not to use AI for content creation.

What are Best Practices on When and When NOT to Use AI for Content?

This is a tricky question to answer as there is no “one size fits all” approach regarding AI consideration. Much of this is due to the fact that the topic of AI is heavily influenced by feelings from users and those on the receiving end of content on the impacts and implications of using AI. Hence why if you were to search this question on Google, you would get a myriad of responses. Our goal as social media experts is to help simplify the process for you (as much as it can be simplified) as well as set proper guardrails on the use of Artificial Intelligence in content creation.

When to use AI:

• We recommend using AI for Polishing. This category of integration is just open enough to AI to include the simplification of tasks and consolidation of the process, while wary enough of AI to still leave the actual creation of content to your skills.
• Producing AI can be helpful in the behind-the-scenes work. You can use it for brainstorms, idea mockups, and other aspects of content creation that aren’t solely the direct output of the final product. One example of how we have done this is through our work with the power generator company Generac. The photo below was mostly all practical and real. It was shot and lit in a studio, with a real hot dog and props. However, our content producer used Polishing to optimize the mustard writing. After making a handful of good “Gs” with actual mustard at the shoot, she photographed them and used them in a prompt for AI to create a realistic mustard font. After the lettering was auto generated onto the photo, she added shadows and obscured parts of the text manually. The result is great piece of content that melded human work and AI in an optimal manner.

When to avoid AI:

• It’s no secret AI is a very charged topic for most people. For this reason, we recommend avoiding using AI for Producing if possible. The risk -reward ratio for using it to generate content leans toward risky and the potential blowback doesn’t tend to outweigh the good. You don’t have to look hard to find examples on social where audiences respond with heated uproar when they even think content is AI-generated. Also, as good as Generative AI can be with prompts, take this next example as proof that it still has a long way to go. When prompted to see if a video/image Generative AI program could create a “unique hedgehog video game character” it gave us this … I don’t know if I’d say that “unique” should give us Sonic the Hedgehog in a parallel universe. Seems like a lawsuit waiting to happen.

This example highlights just how much Generative AI programs are taking from the internet and other sources. As while we’ve seen services do far better than this, we’ve also seen them do far worse.

What Brands are Using AI Well?

There is a lot of ambiguity these days on whether or not a brand is using AI for their content. It wasn’t too long ago that it was very apparent which brands were using specifically AI-generated videos and images to fill their social pages. The content was anything but subtle. Brands like Coca-Cola dove straight into the trend, not caring if it looked AI-generated and were able to develop a lot of buzz (whether that buzz was good or bad for the brand is debatable). Nowadays, most major brands are likely more intentional with their AI usage. Take Prada as an example. Recently, they sparked conversation by launching a campaign for their new collection that looked like AI-generated photos of celebrities and odd creatures. Sound odd? Take a look at the photos and see for yourself.

Prada has been known to be a disruptor within the fashion world. So, using AI for creative in an industry where craftsmanship is appreciated above all was a bold move. However, it turns out there was no AI-risk taken. Weeks later, to many people’s surprise, the brand announced that the campaign had been done with a visual effects artist, and the only use of AI was to enhance the final deliverables. They received a mixed bag of strong opinions as a result on social. While it was a major risk for a social campaign, it worked well in the end because it made sense for a brand accustomed to taking big swings. Prada was able to prove how it polished more than produced and used the buzz they created to spread the brand name and reputation.

I use this example to highlight again how tricky this topic is right now on social media. There is plenty of AI usage happening currently, but the people and brands doing it best are the ones that are intentional with it, using it sparingly to optimize content and campaigns rather than create them in their entirety. This could be sneaking a miscellaneous asset into a carousel, replacing a background seamlessly, or even making an image seem like AI when it isn’t. The truth is, the brands that do AI the best are the brands that you don’t think are using AI at all.

If you look up “brands that are using AI on social”, you will get a massive list of almost all major brands. I challenge you to go look for the AI content on their social pages. Can you pick it out?

We can give our guesses, but the reality is that the brands doing it best aren’t going to be obvious (they may not even make it onto a Googled “list”). To us, that means not the rise, not the fall, but the continuation of AI content rests on it being done intentionally and subtly.