AI Needs a Human Touch for DEI Efforts

While AI makes impressive leaps and bounds throughout industries, Kurt Merriweather, co-founder, and vice president of The Diversity Movement, says the tech is “not quite ready for primetime” regarding diversity and inclusion.

Merriweather, a featured “motivational moment” speaker at NC Tech Association’s recent Diversity and Inclusion in Tech Summit in Durham, NC says part of the problem is that AI “is not as human as it should be.”

AI, in its current form, needs human intervention as it tries to tackle real-world human problems. “What does it mean to be human?” Merriweather muses. “[Throughout the NC Tech event], we’ve heard some amazing stories. Imagine trying to put in a prompt into AI to make that happen. The problem with AI is that a lot of experiences have been left out … when we think about what it means to be human. There’s a whole group of people that have been left out of the conversation.”

DEI and AI: Getting the Definitions Right

For Merriweather, before fixing any shortcomings in AI when it comes to diversity and inclusion, we must do a better job understanding what those terms mean. “When we think about these things, people don’t really understand what DEI means and what AI is … When we’re confronted by something we haven’t experienced before, there’s fear.”

Related:SAS DEI Leader on Building a Culture of Belonging

“When we think about DEI, it’s about having the right people, the right policies in place,” he says, while AI is about trying to simulate human intelligence through machine learning. Drawing the distinction between the characteristics of DEI compared to AI will be important as both change and evolve.

Merriweather notes the extreme pace of innovation happening in AI and the quick pace of DEI adoption in the years leading up to ChatGPT’s explosion of popularity. And now, with DEI drawing political backlash, the media pendulum is swinging wildly, he says. “AI is the new hotness; everybody is talking about it. And the challenge is that DEI and AI are opposed to each other right now in terms of the consciousness we have -- we don’t think about those two things at the same time. And what I’m arguing for is that we do need to think about those things at the same time.”

As companies race to adopt generative AI in their enterprise solutions, Merriweather hopes that DEI does not get lost in the discussion and that AI development is done in a way that does not create more bias. “What we’re seeing with these models is that they do not have the full data they need to be accurate, that there’s built-in bias,” he says. “Because the technology is not being built with everyone’s input, there are deficiencies.”

Related:Achieving Diversity: The Role IT Plays in DEI Recruiting

AI: Juggling Power and Responsibility

Merriweather marvels at the possibilities around AI and productivity. But he’s also worried about the potential for misuse. “There’s speed and efficiency -- there’s creativity. We’re using AI to be a brainstorming partner and thinking about ways that we can learn new things quickly … so it’s great and it’s really helpful. The problem is that there are a lot of challenges with how it’s being rolled out.”

And his fears are backed up by numerous studies, including a late-2023 IBM survey of 8,500 IT professional that showed 42% of companies were already using AI screening in recruitment and human resource tasks while another 40% were considering integrating AI in recruiting efforts.

“There has been a gold rush. There’s a lot of investment in AI as a result of where people are throwing that money and not thinking about it.” When it comes to AI, “all perspectives aren’t being included, because the AI developers themselves are not diverse.”