As a roboticist, people expect me to focus on developing algorithms, prototyping new robots, testing ideas and hypotheses in simulations. And I do all of that. But what really excites me is taking a step back and thinking more philosophically about what we’re actually doing. We’re pouring billions into making machines smarter, more connected, more autonomous. Yet we’re still designing them the same way we’ve always done – for us, and only us.
Let me tell the last thing first: this needs to change. Not because it’s philosophically nice, but because our planet literally cannot afford another generation of intelligent machines designed with the same old anthropocentric playbook.
The Pattern We Can’t Seem to Break
From the ancient Greek automatons that poured wine at parties to the latest autonomous vehicles navigating our cities, they all share one thing: they were designed to serve human needs. Period. The Antikythera mechanism tracked celestial movements for human calendars. Medieval automatons impressed royal courts. Modern AI optimizes human productivity. I co-authored an article recently exploring this persistent pattern, and honestly, writing it was both enlightening and frustrating. Enlightening because tracing this anthropocentric thread through millennia of engineering made me realize how deeply ingrained it is in our discipline. Frustrating because I kept thinking:
“We know better now. Why are we still doing this?”
Here’s where it gets interesting – and urgent. The machines we’re building now aren’t like the ones from even twenty years ago. Today’s intelligent machines are connected, autonomous, and genuinely intelligent in ways that would have seemed like science fiction. Collaborative robots. Autonomous vehicles. Smart cities. These aren’t tools anymore; they’re becoming integrated into ecosystems – both artificial and natural. And we’re still designing them primarily to maximize human performance and productivity. The problem? Our planet is sending us increasingly urgent signals that this approach isn’t sustainable. Climate change, biodiversity loss, resource depletion – these aren’t abstract future problems. They’re happening now. And every new generation of intelligent machines designed solely around human needs adds pressure to already strained ecosystems.
The Ecocentric Alternative
So what’s the alternative? I’m proposing, along with my co-authors Joshua Gellers and Henrik Sætra, that we need to reimagine intelligent machines along an anthropocentric-ecocentric continuum. Not abandoning human needs entirely (that would be neither realistic nor desirable), but fundamentally shifting our priorities.
What would an ecocentric intelligent machine look like?
Imagine an autonomous vehicle that doesn’t just optimize for the fastest route, but considers wildlife corridors, noise pollution impacts on bird migration, and air quality in residential areas. Or a smart city system that prioritizes ecosystem health alongside human convenience, managing water flows to support both urban needs and river ecosystems, balancing lighting needs with impacts on nocturnal species. This isn’t about making machines “nice to nature” as an afterthought. It’s about fundamentally rethinking what we’re optimizing for from the very beginning of the design process.
Four Strategies for Change
In our article, we outlined four key strategies for this transition. I believe these are worth discussing because they’re not just theoretical – they’re actionable starting points for engineers working today:
1. From problem-solving to question-seeking
We’ve trained generations of engineers to be brilliant problem-solvers. But who’s defining the problems? Often, we jump straight to solutions without questioning whether we’re solving the right problem. What if we spent more time asking: “Should we build this? What ecological impacts will this have? Are we optimizing for the right outcomes?”
2. From networks to ecosystems
We talk about networks of machines, systems of systems. But ecosystems operate differently than networks – they’re about symbiotic relationships, circular flows, and long-term resilience. Our intelligent machines need to be designed as participants in ecosystems, not as external networks imposed upon them.
3. From exploitation to exploration
Current machine design often focuses on exploiting known resources and optimizing existing processes. Moving toward exploration means being open to discovering new approaches that we haven’t considered – approaches that might prioritize ecological health differently.
4. From monodisciplinarity to transdisciplinarity
This one is close to my heart. Engineers can’t do this alone. We need ecologists, ethicists, social scientists, and local communities at the design table from day one. Not as consultants brought in to fix problems later, but as equal partners in defining what we’re building and why.
I want to be clear about something; I know that this transition isn’t easy. Industries are under pressure to deliver results quickly. Time and resource constraints are real. Existing investments and infrastructure create path dependencies that are hard to break. But here’s what I’ve learned from my research and from talking to engineers around the world: many of us actually want to design more sustainably. We’re just not sure how, or we feel trapped by the systems we work within. That’s why I think it’s crucial to start having these conversations openly, to share strategies and examples, and to support each other in pushing for change. The engineering community needs to own this transition.
What This Means for You
- If you’re an engineer, researcher, or designer working on intelligent systems, I encourage you to ask yourself: Where do my designs fall on the anthropocentric-ecocentric continuum? What would it take to shift them even slightly toward greater ecological consideration?
- If you’re an educator, how are we teaching the next generation? Are we still training problem-solvers, or are we developing question-seekers who understand their responsibility to the broader ecosystem?
- And if you’re simply someone who cares about technology and the environment, start asking questions about the machines being deployed in your community. Who benefits? What are the ecological costs? Could this be designed differently?
I genuinely believe we’re at a unique moment. The intelligent machines we design in the next decade will shape our relationship with the planet for generations. We have the technical capability to build extraordinary systems. The question is whether we have the wisdom and courage to build them differently. This isn’t about being anti-technology or anti-human. It’s about caring for all species, for ecosystems, for the planet that sustains us all.
If you’re working on these questions or want to discuss strategies for more ecocentric machine design, I’d love to hear from you. This is a conversation we need to have as a community, and every perspective matters.
Didem Gürdür Broo 