I’ve been following the story of Carolyn Marvin’s arrest in Philadelphia, and it’s stuck with me in a way most tech news doesn’t. She was charged with disorderly conduct for blocking a delivery robot’s path, refusing to move until police arrived and arrested her. Her stated reason: these robots are taking jobs from human delivery workers.
On the surface, this looks like a straightforward clash between progress and protest. But dig deeper, and you’ll find something more uncomfortable: a mirror reflecting back at those of us building these systems.
The Comfort of Dismissal
It’s easy for technologists to dismiss Marvin as a Luddite or someone opposed to progress. We’ve heard these arguments before. Automation has disrupted labor for centuries, and society has adapted. The printing press displaced scribes. Factories displaced craftspeople. The internet displaced… well, a lot of things.
But there’s something different about the current wave of AI and robotics. It’s not just the speed of displacement. It’s that we, the builders, have largely absolved ourselves of responsibility for what happens after deployment. We ship the product. We optimize for efficiency. We celebrate the metrics. And then we move on to the next thing.
Marvin worked in delivery. Maybe she still does. Either way, she saw her industry being fundamentally restructed by technology created by people like me, people like many of you reading this. And she decided to protest in the most direct way she knew how: by standing in the way.
Did she accomplish anything? No. The robots are still rolling through Philadelphia. But she did something more important: she made the invisible visible. She forced a moment where we couldn’t just abstract away the human cost into quarterly metrics and efficiency gains.
What We’re Not Talking About
I spend a lot of time in developer spaces, and I notice what we don’t discuss. We celebrate when an AI system can do 100 jobs worth of work. We talk about the technical elegance of the solution. We rarely talk about the 100 people who now need to find something else to do.
I’m not arguing against automation. I’m arguing for honesty. The delivery robots in Philadelphia are solving a real problem: last-mile delivery is expensive and difficult. They work 24/7. They don’t get tired or sick. They make economic sense for the companies deploying them.
What doesn’t make economic sense, under current incentives, is planning for the workers those robots replace. That’s not a technical problem. That’s a business and policy problem. But it’s a problem we create by building the technology in the first place.
This ties directly to broader concerns about AI ethics and responsibility. When we deploy autonomous systems at scale, we’re making choices about how society operates. Those choices have winners and losers. The honest question is: should the builders have any stake in managing the fallout?
The Developer’s Dilemma
Here’s what I wrestle with: I don’t think the answer is “don’t build.” The robots work. They’re cheaper and more efficient. But I do think the answer is “build responsibly,” and that phrase means something concrete.
It means thinking about downstream effects before they become someone’s unemployment. It means advocating for policy changes that make it easier for displaced workers to transition. It means, at minimum, being honest about the costs of the system we’re creating.
When companies deploy autonomous systems at scale, they should be required to contribute to transition support for affected workers. Not as charity. As part of the legitimate cost of doing business.
Marvin’s arrest probably changes nothing about the robotics industry’s trajectory. But it should change how we think about our role in it. We’re not just building cool tech. We’re reshaping labor markets, communities, and lives.
The question isn’t whether Carolyn Marvin was right to block a robot. The question is whether we’re willing to have honest conversations about what we owe the people whose livelihoods we’re casually optimizing away.