AI: A path to innovation or dystopia? 

Outlaws, car chases, desolate terrain. When we talk about a dystopian future brought about by AI, most people imagine a scene from Mad Max. But the reality may end up being more mundane. 

Imagine browsing the cookie aisle at your grocery store, searching for something to satisfy your sweet tooth. As you walk up and down the row, variations on a theme emerge: chocolate chip, mint chip, double-chocolate-mint chip…but where’s the shortbread, oatmeal-raisin, or your favorite sandwich-cremes? 

As guardrails are developed to mitigate the larger existential risks associated with AI, it may be how the technology is applied to everyday business practices like product innovation that we need to watch out for. If used improperly—such as forsaking foundational research and industry expertise—the actual dystopian future we might be destined for is wall-to-wall chocolate chip cookies.

CPG innovation: Beyond the tool

As the leader of an organization that specializes in new product innovation, I’m frequently asked whether generative AI is the future of our industry. The answer is complex. Broadly speaking, it’s certainly on track to disrupt the way we innovate—its ability to canvas and synthesize information, generating infinite combinations of product ideas, attributes, or even usage occasions, is already helping manufacturers jumpstart the creative process. It can be used to brainstorm new concepts, expand on ideas, and fine-tune product details and messaging. 

When Wharton School professors recently pit man against machine (i.e., MBA students against ChatGPT) to generate successful new product ideas, the machine won by a landslide. While impressive, it’s important to remember that ChatGPT is synthesizing information that already exists. On its own, it doesn’t necessarily create truly novel ideas. 

And this is where generative AI’s disruptive capacity becomes murkier: What is often lost in the hype is that large language models (LLMs) are only as good as the data they’re trained on and with prompts that are refined enough to elicit useful ideas. It’s not the tool itself that will unlock true breakthrough innovations—it’s how it’s being used.

Unlocking innovation through AI 

So, how can innovators avoid ushering in a future of nearly identical, AI-generated cookie selections? By leaning on foundational research and honing their prompts to play up their strengths and differentiate from competitors.  

Suppose you’re a baked goods company whose brand stands for indulgence, and you have primary research on wine and chocolate pairing trends. You might design prompts inclusive of that research, as well as specific category knowledge, your brand strengths, and organizational goals to help guide the LLM with idea generation—resulting in “a thin, indulgent, dark-chocolate-coated shortbread that pairs well with Port wine.”  

Breakthrough innovations often sit at the intersection of great ideas colliding across categories—like the iPhone, a revolutionary combination of computer, phone, camera, and MP3 player that changed how we communicate. It’s these moments of disruption that have the potential to transform entire industries and deliver value to consumers. 

When used effectively, generative AI has the potential to accelerate innovation cycles, pulling from a vast repository of cross-category insights that help manufacturers come up with their next breakthrough even faster. This could lead to the rapid development of transformational products that we can’t even fathom today. 

I believe future breakthrough innovators will be those who leverage generative AI tools in conjunction with innovation expertise and unique consumer insights to create the next game-changer in their industry. And I can’t wait to see what the future holds.