What I Learned About Bridging the Sociology Gap Inside AI Companies

Jan de Vries · 2026-07-23

When I first read Dr. Stephen Whitehead's piece on closing the sociology gap inside AI companies, it hit me hard. He's not just talking about hiring a few sociologists and calling it a day. He's talking about fundamentally reshaping how we build technology. And honestly, that's something we've been missing for way too long. AI systems are only as good as the data they're trained on, and that data comes from us—flawed, biased, messy humans. If you don't have someone on your team who understands the social context behind that data, you're basically flying blind. Let me break down why this matters and how you can actually make it work. ### Why Sociological Oversight Is a Game Changer Here's the thing: most AI companies are run by engineers and data scientists. They're brilliant at math, algorithms, and optimization. But they often miss the subtle ways that societal structures, cultural norms, and historical inequalities sneak into their models. That's where sociologists come in. Dr. Whitehead argues that sociologists can help identify bias before it becomes a PR nightmare. Think about it: a facial recognition system that works poorly on darker skin tones isn't just a technical glitch. It's a reflection of who was in the training dataset—and who wasn't. A sociologist would catch that early. - They spot hidden assumptions in data collection. - They flag how certain groups might be underrepresented. - They ask questions like "Who benefits from this feature?" and "Who gets left out?" This isn't about being politically correct. It's about building products that actually work for everyone. And in a market where trust is currency, that's a competitive advantage. ### How to Bridge the Gap in Your Own Company So how do you actually close this sociology gap? It's not as hard as you think, but it does require intentionality. Here's what I'd recommend based on what I've seen work in the industry. **Start with hiring, but don't stop there.** You don't need a full sociology department. Even one person with a background in social sciences can shift the conversation. They can sit in on product meetings, review datasets, and challenge assumptions. Make sure they have a seat at the table—not just as a consultant, but as a core team member. **Build feedback loops.** Create a process where sociological insights are actually used. For example, before launching a new feature, run it by your sociologist. Ask them: "What could go wrong here from a social perspective?" Then act on their feedback. It's that simple. **Train your engineers.** You don't need everyone to become a sociologist, but basic awareness goes a long way. Host workshops where engineers learn about implicit bias, structural inequality, and how their code impacts real people. Make it practical, not preachy. ### A Real-World Example Let me give you a concrete example. A friend of mine works at a startup that builds AI tools for hiring. They were using a model to screen resumes, but it kept favoring candidates from certain universities. Turns out, the training data was full of resumes from people who went to top-tier schools—which correlated with socioeconomic status, not job performance. A sociologist on the team pointed out that this was replicating class bias. They adjusted the model to focus on skills and experience instead. Result? They hired a more diverse team and actually saw better performance. That's the power of sociological oversight. ### The Bottom Line Closing the sociology gap isn't just a nice-to-have. It's essential if you want to build AI that's reliable, inclusive, and trustworthy. Dr. Whitehead's vision is spot-on: we need more voices in the room, especially ones that understand the social fabric of the world our technology operates in. So if you're building an AI company, take a hard look at your team. Do you have someone who can spot the blind spots? If not, it's time to make a change. Your users—and your bottom line—will thank you. This post originally appeared on The European Magazine.