Rodney Brooks on the Robotics Renaissance: Beyond the Hype to Human-Centric Machines

Why a Robotics Pioneer Says We’re Chasing the Wrong Future

It’s easy to get swept up in the hype. Videos of humanoid robots folding laundry flood our feeds, CEOs promise trillion-dollar markets, and venture capital flows like water. It feels like a science fiction future is just around the corner. But what if the field is sprinting in the wrong direction?

Rodney Brooks, a foundational figure in modern robotics, isn’t just skeptical, he’s issuing a wake-up call. The co-founder of iRobot (creator of the Roomba) and former director of MIT’s AI lab argues that robotics has lost its way, seduced by flashy demonstrations and biological mimicry instead of solving real human problems. He sees billions being poured into “pure fantasy thinking” while simpler, more reliable, and more collaborative technologies are overlooked.

This isn’t the grumbling of a techno-pessimist. It’s a course correction from someone who has spent 50 years in the trenches, building robots that work in homes, disaster zones, and factories. He believes we need to stop asking “Can we build a human?” and start asking “What useful work can a machine reliably do?” The answer, he insists, will look nothing like a sci-fi android.

The Core of Brooks’ Critique: Three Flawed Assumptions

Brooks’ criticism isn’t a blanket dismissal of progress. It’s a targeted dismantling of what he sees as the field’s misguided foundations. His argument rests on challenging three pervasive assumptions.

1. The Myth of Imminent Human-Level Dexterity

The central promise of humanoid robots is that they’ll manipulate the world as skillfully as we do. Brooks calls this the current field’s “raison d’être”. Yet, he points out that after 65 years of research, true dexterity remains a distant dream.

The problem is both physical and computational. Our hands are packed with about 17,000 specialized touch receptors, allowing for exquisite force feedback and adjustment. Modern robots lack this. Most industrial robots still use simple suction cups or parallel jaw grippers, technology that’s half a century old. While labs show impressive videos of multi-fingered hands, these demos rarely generalize beyond a single, carefully orchestrated task.

Brooks is particularly skeptical of attempts to teach dexterity by having robots watch video demonstrations. He argues this approach misses the most critical data: the feel of force and contact during manipulation. As he wryly notes, the shaky, primitive pinching grasps in many “humanoid theater” videos tell the real story.

2. The Safety Problem We’re Not Talking About

Humanoid robots are often marketed as collaborative, designed to work safely alongside people without cages. But Brooks highlights a dangerous reality: bipedal robots are inherently unstable and potentially hazardous.

They use walking algorithms that, when they start to fall, pump enormous energy into their joints to recover. If they fail, all that kinetic energy is released dangerously. A robot twice the size of today’s models would pack eight times the harmful energy. This is why, as Brooks observes, companies demoing humanoids often tell people to “keep away” while the robot is walking, a stark contradiction to the promise of seamless collaboration.

3. The Misplaced Focus on Form Over Function

There’s an unexamined belief that because our world is built for humans, the best robot must be human-shaped. Brooks urges us to question this. The goal should be effective function, not biological mimicry.

He predicts that in 15 years, successful robots doing human-like jobs will have wheels for stable mobility, multiple arms for parallel tasking, and specialized sensors, looking nothing like a human. The form should follow a clear, deliverable function. His Roomba is a perfect example: its low, disk-like shape promises it can clean under furniture, and that’s exactly what it does, no more, no less.

Brooks’ Blueprint: The Three Laws for Real-World Robots

Beyond criticism, Brooks offers a constructive framework for building robots that actually succeed outside the lab. He codifies this in his “Three Laws of Robotics,” a pragmatic guide for designers and entrepreneurs.

• Law 1: The Promise of Appearance. A robot’s design tells users what it can do. It must deliver on, or slightly over-deliver on, that promise. A robot that looks capable but constantly fails destroys trust and adoption.
• Law 2: Preserve Human Agency. Robots must not make people’s jobs or lives harder. They should never block, hinder, or remove control, especially when they fail. He points to delivery robots in hospitals that block corridors during emergencies as a prime example of what not to do.
• Law 3: The Long Road to Reliability. Technologies need 10+ years of steady improvement beyond lab demos to become cheap and reliable enough (99.9% uptime) for real-world use. Every additional decade adds another “9” to that reliability. Real-world deployment is an entirely different, harder challenge than a staged demo.

The Industry Hype vs. The Pragmatic Path

The contrast between Brooks’ vision and current trends is stark. The International Federation of Robotics identifies top trends for 2025, including AI integration, humanoids, and sustainability. Meanwhile, firms like Tesla and Figure aim to build general-purpose humanoids, with Figure recently valued at a stunning $39 billion.

Brooks doesn’t deny AI’s role but cautions against overestimation. He calls large language models (LLMs) “masterful bullshitters” because they generate plausible text without understanding truth or physical reality. For robots acting in the real world, this lack of grounding is a fundamental limit.

He also identifies a social pressure he calls “FOBAWTPALSL” (Fear of Being a Wimpy Techno-Pessimist and Looking Stupid Later). This fear, he argues, stops people from critically examining hyperbolic claims, creating a herd mentality where everyone flocks to the trendiest idea like “5-year-olds playing soccer”.

A More Human-Centric Future: Collaboration, Not Replacement

So, what is the right path? Brooks’ career provides the model: build machines that assist and augment humans, not replace them.

• Focus on Collaboration: His robots at Rethink Robotics were designed to be interrupted; a worker could simply grab Baxter’s arm and move it. This preserves human control.
• Solve Specific Problems: The Roomba solved a tedious chore. The Packbot was built for hazardous terrain and was battle-tested in war zones and nuclear disasters like Fukushima. Each had a clear, valuable job.
• Respect the Adoption Curve: Technological change in the physical world is slow. Brooks notes that even digital transitions, like moving the internet from IPv4 to IPv6, take decades. We must be patient.

Getting Robotics Back on Track

Rodney Brooks’ message is ultimately one of optimism, but it’s optimism grounded in reality. The potential for robotics to improve our lives, by taking on dull, dirty, or dangerous work, is immense. But to realize that potential, the field must reorient itself.

It must move beyond “humanoid theater” and the pursuit of artificial humans. Instead, the focus should be on creating quietly competent machines that solve well-defined problems, work reliably alongside people, and respect human primacy. The future of robotics isn’t about building our rivals. It’s about building our best tools.

What do you think? Are we over-investing in biological mimicry, or is the humanoid form the inevitable path to a general-purpose machine? Share your perspective on the future of robotics in the comments below.