A Conversation with Michael Rivera, Executive Creative Director at Upshot Agency
By R. Larsson, Advertising Week
“The genie is out of the bottle,” Michael Rivera, Executive Creative Director, Upshot Agency says. “We should only expect the AI genie to continue improving and becoming more powerful in its capabilities over time. That’s what it’s designed to do.”
Given that, where do we draw the ethical divide between humans and machines in the future of the creative industry? And are creatives in danger of losing their jobs?
Advertising Week 360 spoke with Rivera about this and more, including his more hopeful, glass-half-full outcome for the future of creativity in the age of AI.
Q: ‘AI and Creativity’’ means what? Can you elaborate on what is currently taking place in the creative world regarding AI or is this still theoretical?
MR: AI and Creativity are converging in a very real way, challenging every aspect of creativity from conceptual to craft. It is a moment to be sure but one without much understanding of how generative AI can or should be applied to and utilized in the creative industry. AI is definitely no longer theoretical in our space. It most certainly changes how we define creativity and how we generate ideas and explore ideas. Generative AI is forcing us to evaluate how we use AI in creative and responsible ways and how creativity is empowered and possibly compromised by AI. AI is currently a toy to be explored, played with and tested in the creative world. It is a toy in the toybox, not yet a tool in the creative toolbox. That said, it is conceivable that generative AI can be a useful tool as we come to better understand how we can use it, how we should use it and maybe even when we should – and shouldn’t use it.
Q: What do you mean?
MR: Unlike creative toolbox innovations of the past, such as the typewriter, the desktop computer or creative design apps, these AI systems don’t just help to enhance an idea or speed up the expression of the idea, they also assume the role to some extent of coming up with an idea. AI systems, even at this early stage, can already make myriad creative decisions typical to the human creative process in only a few seconds. How much is being lost in the creative process when crafting and creating connections among images, colors, sounds and words are optimized and streamlined for efficiency and human-like recognizability? There is something fundamental missing when the imperfect, inefficient process of constructing human ideas is optimized. Where is the creative integrity when images and words and design elements are commoditized as assets to be grabbed and assembled by a system engineered to assemble an answer passably recognizable to humans?
Q: You’re making a distinction between AI for creativity or iteration I believe?
MR: Yes. The learning nature of generative AI is designed first and foremost for efficiency and scale. Is that creativity or just relentless iteration? Maybe we’re not at a crossroads of AI and Creativity, where there’s a clear either-or choice. Instead, it could be we’re at an on-ramp to a truly evolved creative industry where the efficiency and speed of AI, when used with skill and true creative talent, allows creators to do more, explore more and learn more than ever before in human history.
Q: So if AI is to play a bigger role in the creative process, where is the line drawn between artist and machine?
MR: At the moment, it seems that there are two potentially impactful roles for AI to play in the creative process. One will likely be a tool in early-stage ideation. AI could serve as a sort of low-rez, high-volume idea generator that could provide unexpected – and probably lots of expected starting points for how to think about a creative challenge. Another role for generative AI could be further downstream in the creative process as a powerful iterator of creative executions across media and experience environments in a unified experiential way, beyond what’s already been used to optimize basic images and copy. It’s also worth considering how active AI could play a role in brand experience.
Q: Can you provide an example?
MR: Imagine an AI system in a branded environment with an invitation for the audience to use the brand name as part of a prompt to generate memes, mash-ups and content. The line between artist and machine exists in the space between the artist’s human inspiration and the AI system’s iteration of those prompts. For now, without a human-inspired idea or command – not to mention the ocean of human-generated data that the AI system learns from via the internet – the AI system effectively has no proactive role in the creative process other than a powerful ability to respond with speed and efficiency.
That could change as AI gets smarter and could possibly evolve to understand and react to something as broadly complex as deciphering a marketing challenge and then authoring an insight, a creative brief, resulting in the creation of a humanly-nuanced and relevant campaign of integrated creative solutions across strategically sound digital and real-world media and experience environments. Until then, it seems humans are still the true artists, the organic creatives whose role it is to create the sparks of inspiration that ignite the creative process and connect ideas in all their forms.
Q: Ethically speaking, do you see a problem with integrating AI in the creative space?
MR: There are many ethical considerations and concerns beyond those that might be in the creative space – including information security, identity, gender bias and management of misinformation, among others. As for the ethics of integrating AI in the creative space, we’re at the early stages of the discussion. There are some obvious questions and concerns that seem relevant immediately.
Such as how do we ethically allow for the correct attribution of creative elements used by AI, such as brand assets, celebrity and influencer likeness (image, voice, and presence) and published music and language? Also, should there be specific identification of who the human creators are and were they AI-assisted? Who actually created what? Was the work created by AI and humans working together? Or mainly AI and minimal human authorship? Or nominal initial AI involvement and mostly human crafting in the execution? Those three examples of integrating AI into the creative space only scratch the surface of the nuanced ways we will use AI in the creative process, never mind the debates we can expect around the ethics of doing so.
Q: If we are to believe AI will only improve, are creatives and designers in trouble or losing their jobs?
MR: The genie is out of the bottle – and we can expect that AI genie to only continue to improve and become more powerful in its capabilities. That’s what it’s designed to do. As a result, there are two things that will matter most to the careers of creatives and designers in an AI-enabled creative world. The first thing that matters is the creatives’ willingness to accept AI as part of a contemporary creative process.
Today it’s very much a toy. Soon it will be a legitimate creative tool, to be mastered and applied skillfully and ethically. The second thing that’s important to creatives will be their ability to adapt AI in the best ways possible to enhance and elevate their work –their creativity. The humanity of art will prevail, I believe, as long as humans continue to create with a sense of purpose and craft. Everything humans create, publish and post ultimately feeds the brain that is generative AI. We exist without AI. AI does not exist without us.
Q: Glass half-full, how can AI help improve creative thinking?
MR: There are basically endless possibilities. That’s a key feature of the system. It’s a spark generator. An additional perspective. A creative partner who will show you drafts of ideas that you could refine, reassemble or reject and learn from. AI depends on human expression as fuel for its engine. Human creative thinking plays a unique and indispensable role in the development of artificial creative thinking.
Q: Alphabet’s DeepMind features an AI tool for screenwriters that can generate character descriptions, plot points, location descriptions and dialogue. If they can do it for screenplays, a thirty-second ad can’t be far behind, right?
MR: You could argue that many thirty-second ads are already painfully formulaic. However, a film script is a work of art, whether it’s good or bad being beside the point, that primarily serves the story that the artist intends to tell. The film script has few objectives outside of expressing the artist’s vision.
The objectives of the thirty-second ad are substantially different from that of a film script. Writers of ads should of course aspire to create a form of art but also they must accommodate many more objectives, selling things among them. Ads must drive commerce, they have to engage, inform, entertain and ultimately drive people to buy things, not just ideas. It’s not enough for an ad to exist only to be interpreted. Ads have to connect with specific groups of customers or potential customers. They have to resonate in a way that inspires action. Even the most esoterically artistic commercial is designed to inspire people to follow a brand to purchase that brand.
To date, it seems most generative AI commercials are devoid of the human touch and are the result of creating scripts that sell an assemblage of the information keyed in as prompts. Not emotionally resonant. Not insightful. Just basic. For example, the recent Ryan Reynolds Mint Mobile spot, which is an AI creation, falls flat despite being delivered by one of the most entertaining comic actors in the business. Bad creativity? Don’t blame the AI. It’s just doing exactly what it’s designed to deliver – an amalgamated composition, authored as quickly and efficiently as possible. Not bad. But not good either.