Will robots give us more "me" time?

The robots are coming, and Daniela Rus is OK with that. As the director of the Computer Science and Artificial Intelligence Laboratory at MIT and a MacArthur fellow, Rus has been at the forefront of robotics and AI for decades. Her new book The Heart and the Chip, co-authored with science writer Gregory Mone, argues for a vision of AI and robotics that prioritizes people: “Robots can be designed to filter the repetitive tasks out of our daily lives,” she writes. “By working closely with intelligent machines, we could have more time to be human.” 

The book chronicles a career highlighted by human-focused achievements in robotics. In one, her team experiments with using autonomous car tech to help blind people navigate their environment on foot. In another, Rus helps develop a robotic pill designed to remove the thousands of button batteries ingested by children each year without the need for surgery.

The real fun is in the back half of the book, where Rus explains some of the nitty gritty behind the field’s challenges: why it's incredibly hard to engineer a robot that can change a light bulb; why roboticists often dread giving live demos; and why building a robot that could clear dishes off of a dining table would be more of a feat than sending one to Mars (robots love avoiding contact, you learn.) The deep dive on these obstacles underscores Rus's belief that overcoming them could transform our lives.

In an era where time is increasingly precious, Rus argues that robotics and AI are keys to unlocking a more efficient, creative, and fundamentally human existence. Here, she explains her vision:

One thing that struck me in the book was your assertion that “anything can be a robot.”

Most people think of metallic humanoids when they hear the word robot. But robots don't have to be inspired by the human form. They don't have to be boxes on wheels. We’ve been rethinking what a robot is. In my lab at MIT, we've created soft robots that look like fish and sea turtles, and these robots help us understand the mysteries of the ocean. We've also built micro-robots, like the ingestible robotic pill that's made out of food. This type of robot will be able to deliver incision-free surgery. My colleagues are even designing robots out of biological cells.

You write about “soft” robots and wanting to design one for your aging father to regain strength and mobility. How do you see robotics augmenting our physical bodies?

We are working on robots that can be partners to people, to restore or enhance our human capabilities. An exoskeleton is a wearable device that acts as an external framework or structure designed to support and augment a person's movements, strength, or endurance. Traditionally, exoskeletons were large and bulky, and constructed from metals and hard plastics. They were not very comfortable. But bringing materials and machines closer together, with materials becoming more and more intelligent, and machines becoming more and more like materials, is enabling us to really start to dream up robotic clothes or exoskeletons that could conform to the natural movement of the body. In the future, your jacket or pants could be robots.

Where does AI stop and robotics start?

Today, AI and robotics are largely separate fields. But this is starting to change. This is because AI is about to break free from the computer screen as we begin to fuse the digital intelligence of AI with the mechanical prowess of robots. In doing that, we will achieve what I like to refer to as physical intelligence. This is where AI, machine learning, and robotics will start to open so many opportunities for us. Robots will be taking on more routine tasks so that people can focus on the more critical-thinking and creative aspects of their jobs and their lives. 

Speaking of creativity, I wonder if at least some of your optimism comes from the fact that you still see a pretty big gulf between AI and humans. You write that AI is never going to produce a Shakespeare or Tolstoy, for example. 

I think of AI as a tool. It could help us be better artists by creating with different media. Consider the artist Refik Anadol. He views AI algorithms as the new paint brush, and the data as the new paint. His installations are mesmerizing. There's one in the lobby of the MoMA in New York. Refik is an example of an artist using AI to enhance his own creativity.

To me, great art captures something profound about the human condition, or makes a sort of social commentary. It offers insights and reflections that are deeply rooted in the personal and collective experience. It’s human driven and human expressed. AI can help us with our expression. But AI does not replace the artist.

You write about some of the medical use cases of AI—teaming up with doctors to identify cancer, for one. Do you see a similar kind of partnership happening in other fields? 

In the Harvard project you’re referring to, doctors and machine learning systems were given scans of lymph node cells, and they were asked to label the scans “cancer” or “not cancer.” Doctors working alone had a 3.5% error rate, and the machines had a 7.5% error rate. But when they worked together, the error rate dropped to 0.5%. This is because humans are better at some things, and AI tools are better at other things. And they could get better outcomes together.

We are going to see AI team up with humans across professions. Give me a job and I will tell you how AI could act as a copilot: teaching, lawyering, finance. Everything.

But one thing to keep in mind is that because today's AI algorithms are not perfect, we have to stay in charge of making the decisions. I like to think of AI as an intern that runs around, pores over data, and then brings me interesting insights and suggestions. I then use my own expertise and knowledge to evaluate them and come up with an action that’s consistent with my objectives and ethics.

As we rely more on AI and robotics, are you concerned at all that our skills and creativity might atrophy?

Broad education continues to be important. Right now, it is also very important to teach ourselves about the capabilities and limitations of the AI tools. 

We will also need a different approach to how we learn. Right now, we learn and work sequentially. We go to school, study, and at some point, we're done studying and we enter the workforce. But the future will require that education continues throughout our lifetime…because there will continue to be new tools that will be transformational, and it will be very important to keep up with them. 

At the same time, it's also important to cultivate an environment that values creativity, critical thinking, collaboration, communication, problem solving—essentially human qualities.

You write about your very real need for an AI that answers your emails. Compared to some of the other robotic projects you’ve been a part of, that seems… doable?

My mythical AI tool is a personal assistant that helps me optimize all aspects of my personal life and work life in order to ensure that I live well and work effectively. It doesn’t exist yet. 

I get hundreds of emails every day. I would love to have an assistant that learns about my approach to solving problems and my communication style from the hundreds of thousands of emails that I’ve sent. When a new email comes in, I don't want AI replying on my behalf. But I would like a first draft that I could look at and edit as needed. The more predictive capabilities we develop, the better the tools will be at anticipating the needs and making recommendations proactively. I also think that these tools will get really great at personalization.

This takes me back to this idea that I shared with you earlier, that AI should be like an assistant that can do things to alleviate the volumes of work we have to take care of every day. 

With all this time thinking about robotics and AI, what does being human mean to you now?

To me, being human is everything in the human experience that technology cannot replicate. We have emotional depth, we have creativity, we have empathy, we have organizational and social structures, we are capable of advanced tool development. And we have ethical judgment. 

And when I consider AI, I really am truly amazed by the humans who are creating it. We’re the only species capable of building these extraordinary tools. But that also comes with a lot of responsibility to ensure that the tools can be used for the greater good, and they benefit the planet, the plants and animals we share this wonderful place with, and the future generations. I remain convinced that we can do it.  

This interview has been edited and condensed for clarity.