Generative AI: The new “right hand” of creativity

Belén de Areba

26 April 2024

rta creatividad banner gr 1

If someone had told us a few years ago that Tom Cruise would become one of the most famous tiktokers, we would have thought that was nearly impossible.

We live thinking about the future, and sometimes it’s easy to lose sight that the future is already here. The digital revolution is accelerating more and more, leaving us only one alternative: to adapt and learn day by day.

A clear example is the rise of generative Artificial Intelligence. Its speed and independence from external resources have changed the rules of the game, significantly boosting productivity and efficiency.

However, it is precisely this volatility that may cause many marketing departments to hesitate when deciding to opt for this type of tool. Yes, there are many uncertainties, but we cannot afford to wait and see what others do, or else we will have missed the stop. It’s about researching, testing, experimenting, making mistakes… choosing progress over perfection.

A mantra that at Good Rebels Creative we have taken to heart and have tried to incorporate into our entire creative process (in fact, this article could perfectly well have been structured and written by an Artificial Intelligence tool, but I neither confirm nor deny). These are some of the learnings that this process has brought us.

captura de pantalla 2024 04 26 a las 10 03 18

Text Generative AI as an engine for human creativity.

Although there is still debate about whether AI can truly be creative—this is worthy of another article—it has been shown that ChatGPT4 can be more creative than some university students, and an image generated by AI won in the creative photography category at the Sony World Photography Awards. Its true value, however, lies in its ability to enhance human creativity.

For example, it can boost writing and creative ideation processes. According to a recent experiment with 453 professionals who regularly generate texts as an integral part of their work, including marketing specialists, the use of ChatGPT increased the quality and originality of texts by 18%.

The important thing is to understand that these tools are a resource to help us perform certain tasks, but the final word will always be that of the professional. For example, in the following scenario, the response provided by ChatGPT is still very standard, and we would not use any of the names it suggests. However, we can draw from five new paths to continue thinking and come up with the final proposal.

captura de pantalla 2024 04 24 a las 17 03 03

In short, the value provided by a creative professional is not only linked to the deliverable; we dedicate a large part of our time to the process. And it is in this process where our true contribution lies: in the ability to think strategically, to interpret and connect ideas, and to shape innovative and effective solutions. AI can be a powerful tool in this process, but creative spark and strategic vision remain irreplaceable by artificial intelligence.

Image Generative AI: a matter of prompting and timing.

If generative AI is already the “Roman Empire” of creative professionals, image generative AI is undoubtedly the star subcategory.

However, when it comes to producing image content using these tools, we face two fundamental challenges. On one hand, generating effective prompts that, in a few steps, allow us to obtain the desired result (only then does AI make us truly efficient).

On the other hand, we must understand the users’ willingness towards using AI in our brand content and choose the right consumption moment to introduce such content.

The art of prompting.

“I can’t get it to do what I ask” is one of the most repeated phrases after hitting enter when making a request.

Considering that these programs are constantly learning through algorithms, the solution is clear: test, test, test, and test again. These are the principles we apply at Good Rebels to make the result as accurate as possible:

  1. Define the purpose and objective. We will clearly define in which channel the content will be used, for what type of brand, on what occasion (for example, a product launch), and what we intend to achieve with the action.
  2. Define the target audience we want to reach. Don’t assume that AI knows your audience—remind it!
  3. Choose the personality. We must be clear from which standpoint we want the generator to give us the result, what role we want the generator to adopt to bring it closer to the brand’s typology.
  4. Provide context and relevant information. The more details we provide, both about the art direction and the elements represented in the image, the closer the result will be to our expectations.
  5. Add reference images. We can contribute by showing another image that resembles the expected result, but be careful with commercial use if you start with a copyrighted image.
  6. Choose the action commands. Before formulating our request, we must clearly define what action we want the AI to perform: create, imagine, exemplify, describe, mix…

captura de pantalla 2024 04 26 a las 10 05 48

Furthermore, there is always the possibility of using multiple platforms to achieve the desired goal, as each processes information differently. For example, requesting MidJourney to create a “Disney”-style character will yield different results than doing it through Bing.

captura de pantalla 2024 04 25 a las 9 49 58
MidJourney vs. Bing | Prompt: “A disney pixar girl in town”

Fortunately, tools like Photoshop Beta are already working to make our lives easier by integrating generative AI into their interface, simplifying the prompting process and speeding up results.

captura de pantalla 2024 04 25 a las 9 51 06

AI faces hate too.

Undoubtedly, AI is on the rise, and it’s becoming increasingly common to encounter ‘fake’ images generated by AI on social media (for example, inspiration accounts are filled with them). However, user acceptance is not always positive, and there’s a risk of receiving negative comments for resorting to these tools instead of professional creatives. In this context, a counter movement is emerging, known as the ‘human-made seal’, whereby some brands and content certify that their content has not been generated with AI.

So, how can we minimise risks? The key lies in understanding when AI integration works best for content consumption, and entertainment is undoubtedly one of them. Instead of trying to pass off AI-generated content as real, involve the user in the brand’s decision to use AI, joining trends or creating participation dynamics, such as challenges to differentiate between real and AI-generated photos.

Generative Video AI: The Next Big Thing?

If in 2023 we witnessed the rise of Generative AI in text and image combination, in 2024 all signs point to the great revolution being in the realm of video. The ability to bring images to life, creating out-of-home fakes, and ads crafted with artificial intelligence are opening up a new horizon in the advertising field.

Though we are yet to witness the boom of this technology, there are already several platforms offering sophisticated enough solutions to incorporate them into our creative asset production workflow, such as Runway, which can generate a video from an image, another clip, or even text.

However, undoubtedly, the one that is causing the most buzz is Sora, for the quality, complexity, realism, and high resolution of the videos it generates from just a phrase. We describe what we want to see, and there it is!

captura de pantalla 2024 04 26 a las 10 08 30

Beyond these seemingly science fiction tools, there are other AI-based tools that, while perhaps less surprising at first glance, are making our lives easier and speeding up video generation processes. A clear example is CapCut, one of the most versatile and accessible applications we can find on our mobile devices and can be used daily on our social networks. This application offers a wide variety of predefined templates that allow us to create both static and animated effects, change backgrounds, and add subtitles automatically, as well as providing a wide range of transitions and music to customise our videos.

captura de pantalla 2024 04 26 a las 10 08 38
Image created by AI

 

 

 

 

What about Generative Audio AI?

In a context of digital audio boom, it’s essential to address how AI is transforming this realm, from voice generation to music and sound effects creation. Although this may be an area where we haven’t fully glimpsed the potential of AI yet, it’s undeniable that it’s advancing significantly and will soon be a topic of conversation.

While it’s true that voice generation through AI is quite developed in the Anglo-Saxon world, finding a perfect sound in Spanish, especially with a Spanish accent, can still be challenging. However, tools like ElevenLabs allow us, through trial and error, to experiment with different voices and even clone real voices, adjusting parameters such as accent, speed, vocalisation, gender, or age, aiming to achieve as realistic a result as possible.

Moreover, in recent weeks, we are witnessing the popularisation of generative music AIs, which could greatly simplify the process of creating audiovisual content by freeing brands from depending on stock libraries and enabling them to generate their own sonic identity. TikTok, in fact, is already experimenting with its own functionality.

Will the AI Boom Lead to Infinite Content?

There is no doubt about the many benefits of generative AI and how it can enhance creative work, but we cannot overlook that it’s a new reality with many implications, not only ethical and legal (that’s also worth another article) but also strategic.

At the content strategy level, AI presents us with a double-edged sword: if we take less time to produce, that implies we can generate up to four or five times more content. But, does infinite content really benefit us if user attention levels remain constant?

Responsible use of AI —beyond being attentive to copyright issues, data protection, or reputation— also involves allocating the time we save thanks to it to tasks that generate the most value. Strategic thinking and in-depth analysis of our target audience are good candidates. This way, we can maximise the benefits of generative AI both in the creative and strategic realms, ensuring we use this tool consciously and effectively to improve our work and performance without losing the essence of the brand.