Can we actually build AI that’s both powerful and ethical? Maya Derrick’s piece “Tech & AI LIVE: Are Humans Key When it Comes to Ethical AI?” forces us to face that question head-on. I think what hits hardest here is the realization that AI ethics isn’t some box you tick before launch. It’s messy, human, and demands constant attention.

The practitioners Derrick interviews, Anuj Anand from Ausenco, Rebecca Warren from Eightfold AI, and Rajh Odi from Bristol Myers Squibb all circle back to one difficult reality: technology doesn’t police itself. Humans do. Anand’s point about most AI projects stall at the proof-of-concept stage unless they solve real business problems? That’s the wake-up call. I feel that we’ve been overly swept up in AI hype that we’ve forgotten to ask whether it actually solves anything meaningful.

What really grabbed me was the focus on bias. It seems the answer isn’t just better algorithms but better people like diverse teams who challenge assumptions and spot blind spots before they become disasters. You can’t code your way out of prejudice. You need humans from different backgrounds poking holes in your logic.

Then there’s workforce angle. Warren’s take on “human-centered AI” flips the automation panic on its head. Instead of replacing people, what if AI helped them discover skills they didn’t know they had? Bristol Myers Squibb is doing exactly that, using AI to unlock internal mobility and amplify human potential. I think that’s the difference between ethical AI and corporate cosplay.

Derrick’s article doesn’t offer easy answers, and honestly, that’s refreshing. She makes it clear that ethical AI isn’t a product you ship. It’s a cultural commitment that requires judgment, humility, and the guts to put human values first. Technology won’t save us from our worst instincts. Only we can do that, and it happens one decision at a time.

So here’s the uncomfortable truth: if your AI strategy doesn’t put humans at the center, you’re not building the future. You’re just automating the same old problems with shinier tools. I feel that too many organizations are still treating ethics as an afterthought, a PR spin, something to slap on the website. But the companies that get it right? They’re the ones asking hard questions before writing a single line of code.

The choice is ours. We can keep pretending algorithms are neutral and hope for the best, or we can do the harder thing, build AI with intention, accountability, and actual respect for human dignity. It seems pretty clear which path leads somewhere worth doing.