Listening to Simple Minds’ “Don’t You (Forget About Me)” recently, I was reminded of an overseas trip with my aging parents. The visit brought into sharp focus, for me personally, the impact of relentless technology acceleration, and the obligation to keep pace with broader society’s digital comprehension – as what were once deemed next generation technologies quickly become commonplace.

In my parents’ case, what if their favored method of interacting with a brand or retailer all of a sudden is no longer available, or unrealistic, and replaced by something new? Does that preclude them from accessing services and products that the rest of society can enjoy? To paraphrase Simple Minds, do my parents get forgotten?

Revisiting the definition of digital discrimination
We often think of digital discrimination within the context of AI/ML biases. We might ask why a programmed algorithm takes a specific action – what cultural, societal, or other factors influence the logic which underpins that intelligence?

But there’s another intrinsic (and often invisible) facet to digital discrimination: The unintended exclusion of those who are less digitally proficient from certain services or products, due to the sheer speed at which technology is evolving. Put simply, certain groups are being left behind.

This is the byproduct of technology providers making assumptions on behalf of society – specifically, assumptions which are either not necessarily accurate or are premature and being launched into a market that’s not ready. Diligence is required here, recognizing that while there’ll always be a demand for progress, innovation and evolution, the responsibility to address inclusivity must not be ignored.

Take this very simple example from the UK, where a recent petition attracted over 100,000 signatures from shoppers calling for a major UK supermarket to replace its self-service checkouts with real people. The angst in that petition isn’t directed at generative AI, but it’s a great example of the challenges retailers and brands can encounter when they introduce new digital platforms. Self-checkouts aren’t new – yet still, there’s pushback on their adoption, for reasons ranging from digital literacy through to the desire for human interaction.

Reassuringly, there’s good news. Some brands and retailers are taking note and adjusting their experiences accordingly. Last year, Dutch supermarket Jumbo announced they are opening 200 “chat check-outs” for those who want to actually engage with another human being, as a sizeable cohort its customers may still prefer this type of experience.

So, what does this have to do with generative AI?
Digital discrimination is particularly relevant right now, at the precise moment when we’re being promised that GPT4 and other generative AI models will transform every element of our lives. Generative AI has massive potential to be a positive force for many different parts of society and various groups of people. Whether it’s for accessibility or simply making our daily lives easier, emerging intelligent technologies can (and often do) have a hugely beneficial impact.

Consumer-facing brands and retailers are of course leaping in early, with more intelligent customer interaction among the most prominent immediate use cases being explored. According to data from Coresight Research, 17.8% of the US population already use AI-enabled voice activation to make purchases each week. Coresight expects this will increase to 27.7% by 2025 and 76.0% in 2030. Meanwhile, Carrefour is already experimenting with ChatGPT and generative AI to create content for (and to respond to) customers.

But while this innovation is exciting, retailers and consumer-facing brands rushing to explore use cases around generative AI have a responsibility to ensure they’re not unintentionally alienating certain demographics. As generative AI becomes more mainstream, and human contact is limited to the exception, rather than the norm, there’s a real risk some groups could be left behind.

As with so many things: Balance is key…
… but difficult to strike. Specifically, improving experiences – and making experiences more accessible – without going so far that you risk isolating certain people. And supporting decision-making without “automating humans out of decision-making”. I’ve quoted MIT research scientist Renée Richardson Gosline there, who referenced it when referring to the concept of “good friction” in this HBR article. She describes good friction as “a touch point along the journey to a goal that gives humans the agency and autonomy to improve choice, rather than automating the humans out of decision-making.” To me, that’s a great summary on how we should approach generative AI: as a guiding force, a sidekick or a co-pilot.

Ultimately, for retailers and brands the only way to find the right balance and protect against this unintentional digital discrimination is a deeper, richer and more intricate understanding of your customers. How does a customer segment interact with you now? More importantly, why do they interact with you? In the context of generative AI, where does it make sense to automate – and do you unleash that capability for everyone, or focus on a particular customer segment?

The wisdom to ask why
New technology isn’t going to stop arriving. Nor is the pace of change going to slow down. So, amid this blizzard of evolution, it’s important to acknowledge that technology and digital can sometimes take you farther away from your desired outcome, rather than help better meet customers in the moments that matter. Just because you can, it doesn’t mean you should. Because it may not necessarily be what your customers want or are looking for.

In the generative AI age, having the wisdom to know when to engage, why you’re introducing a new service or solution and what positive impact you can deliver is more important than ever.

It might be the dawn of an exciting new era. But this line is timeless: Don’t you forget about me…

Head here to learn more about Avanade’s approach to generative AI.