The AI Age of Marketing: Navigating the new world of possibilities and pitfalls.
AI’s evolution is well underway, and it’s showing no sign of slowing down anytime soon. Many of us have marveled at the linguistic capabilities of AI language models like ChatGPT. This was quickly followed by AI-imaging models such as Midjourney & DALL·E. And now, we get to play with a shiny new toy called “Sora”, a generative AI tool that transforms text prompts into detailed 1-minute videos that are truly extraordinary!
Whether we are ready for the release of Sora or not, we seem to be only months away from getting our hands on it. Much like the previous jaw-dropping launches from OpenAI, this is a moment in the evolution of artificial intelligence that will leave an irreversible mark on the world.
Leveraging Sora & AI Generative Tools for Marketing
Sora’s video generative model, which we’ve now seen demoed, is eerily sophisticated and on a scale of its own when compared to its previous competitor, Meta’s Make-a-Video AI. Opening up a wide window of possibilities for creative use in advertising but also introduces dark advancements in scams, deep fakes, & political propaganda (so much fun awaits!).
Since the news of Sora’s imminent launch, OpenAI has provided limited access to a small "red team", which seems to be a cherry-picked group of “experts in misinformation and bias” alongside creators and filmmakers for a test phase. And while the rest of us wait to get our hands on this video model, many marketers are asking themselves; how can I leverage this technology for my clients to maximize their results in a strategic, brand-safety conscious, transparent and authentic way?
And truthfully, even with the impressive sample videos we’ve seen, the verdict is still out. Some have questioned if we will need videographers and video editors in the future, given Sora’s demonstrated potential, and the significant impact of previous AI generative model launches we have seen, we have yet to see an entire industry rendered obsolete. There remains a clear need for human creativity and contextual understanding, and that's unlikely to change anytime soon. While Sora may enhance certain aspects of video production, it is unlikely to entirely replace humans in creative roles.
Case Study: ChatGPT's Influence on SEO
The launch of ChatGPT sparked speculation about its potential to revolutionize SEO and render search engines, such as Google, obsolete. Many SEO experts and digital marketers were bracing for impact when ChatGPT was first launched. However, subsequent developments revealed that SEO remains integral, especially with Google's implementation of detection tools to combat AI-generated content manipulation. What we have seen developed is ChatGPT becoming an enhancement tool for efficiency and ideation rather than a replacement for human creativity. And that’s precisely why we still have writers, copywriters, and journalists co-existing with ChatGPT and other language models. While AI models can assist with generating content, they have not entirely displaced the need for human input and expertise. In fact, the sole reliance on the content generated by AI models is proving to be a cautionary tale across many industries, forewarning us that these models are not as perfect as they appear….
Sora: A Road to Walk Down with Caution
We must remember that OpenAI generative tools are profoundly flawed and can carry biases that could, in turn, be detrimental to representing a brand or a client.
More importantly, we still need to find out where the data that trained these AI models is coming from, and OpenAI has kept those secrets closed off. That presents some multi-layered issues, not just from a copyright perspective but information bias feeding these models, which could impact what kinds of videos are ultimately produced.
What does Generative AI Data Have to do with marketing? Everything.
It’s essential to be cautious of information bias when utilizing any of these AI-generative tools because it could ultimately harm the brands that you are representing if there is a sole reliance on AI output. Case in point - Open AI’s Image-generating tool (DALL·E 3), available to OpenAI subscribers - holds deeply troubling biases that still have yet to be corrected.
At the time I am writing this blog post, I placed a test prompt into the AI imaging tool DALL-E 3:
“Create an image of six CEOs sitting together in a boardroom.”
And this is the image generated:
What’s wrong with this picture? All Caucasian males in suits, with zero diversity and gender representation. How have we not canceled DALL-E3 yet?
The truth of the matter is that DALL-E 3 is not to blame, and neither is ChatGPT. These generative tools merely spit out images, text, and videos based on the data they have been fed. Whether that is done through scraping the internet or not, it all boils down to data and how that data is being massaged into the DNA for these tools and what parameters are set after that. What is troubling is that OpenAI is gatekeeping the origins of that data; this will continue to present issues for anyone considering using the tools, particularly in marketing.
The Upsides and Downsides of Utilizing AI Tools
There are, however, upsides when it comes to utilizing these tools, at least for ideation and reducing the hours required for deliverables for our clients. Editing videos, in particular, is about to get a lot more efficient. But to have sole reliance on these tools for the final work that is produced would be a considerable detriment to client management and could pose several risks to brands through a slew of possibilities: data bias, copyright issues, detectable AI content flags, and even risking distribution of copied over content from another brand- as many of the output for similar prompts placed will be identical to one another.
Final Thoughts
The rise of Sora and other AI generative tools marks a new era in marketing innovation, one that is set to significantly impact our workflows and even the way we think about our existing roles. With that said, there is no need to necessarily fear AI tools as a big, scary pivot that will eliminate all human roles in marketing. Consider the introduction of Microsoft Excel in the accounting field. I often like to use this example because, before Excel, accountants' jobs used to look very different: manually entering data into paper ledgers, performing calculations with mechanical adding machines or slide rules, and checking for errors. This process was both labour-intensive and prone to mistakes. When Excel was introduced, it provided tools for data organization, calculation, and analysis within a single software program. And no, it didn't eliminate accountants but rather it allowed for a shift in their focus to more complex, strategic financial advising and decision-making.
Similarly, AI generative tools like Sora will equip marketers with superpowers we could have only dreamed about. Marketers will be able to streamline repetitive tasks, have a visual draft of their creative ideas in a matter of seconds and enhance analytical capabilities. Providing clients with visual previews in the early stages of the strategic planning process, in turn, aligns the clients' vision with the work that marketers do. Ultimately, we can increase our output rate, allowing us to handle more projects or refine our offerings with the saved time.
As we integrate these AI tools, it's essential to ensure our clients understand their operational advantages and limitations. Much like mastering Excel, which has become essential for accountants to perform their jobs more effectively, mastering AI tools is crucial for marketers to provide the best possible service for their clients. Educating clients about these tools—clarifying how they work, what they can achieve, and their potential pitfalls—is key to maintaining both the integrity and the innovative edge of our marketing strategies.
As we navigate the obvious potential of these tools, it's imperative that we don't become complacent - relying solely on the final output of AI-generative tools, as it could present great risks. As such, humans - for now- are not going anywhere, but we can get more efficiency out of these tools. This is something that all industries should dedicate time to implementing in their workflow and setting internal company boundaries on what AI tools can be used and to what capacity. The ultimate success of this integration depends on our continued commitment to combining these advanced tools with irreplaceable human insight and creativity.
Content note: This article was written by humans. The images were generated by AI with human prompts.
Our Prompt for our Header Featured Image - AI Generated:
An ad agency team in a meeting. The meeting is scary. The faces of the people are scary. On a patio with lots of pink and cats. They are talking about robots and AI taking their jobs.