Boston Dynamics AI Institute Targets Basic Research

This morning, Hyundai Motor Group and Boston Dynamics announced the launch of the Boston Dynamics AI Institute, to “spearhead advancements in artificial intelligence and robotics.” BDAII (I guess we’ll have to get used to that acronym!) will be located in Cambridge, Mass., with more than US $400 million of initial investment from Hyundai (Boston Dynamics’ parent company) and BD itself to get things started. Heading up the whole thing will be Boston Dynamics founder Marc Raibert himself, with Al Rizzi (Boston Dynamics’ chief scientist) as chief technology officer.

This new venture looks promising.

Here’s what we know about the Boston Dynamics AI Institute, freshly harvested from their new website and Hyundai’s press release:

We, at the newly formed Boston Dynamics AI Institute, want to accelerate that future, creating future generations of robots that are far better than anything that exists today. Our mission is to solve the most important and fundamental challenges in AI and Robotics to enable future generations of intelligent machines.

We need to make robots smarter, more agile and dexterous, and generally easier to use—more like people. Once we do that, robots and other types of intelligent systems will increase productivity, free people from dangerous work, care for the disabled, and generally help people live better lives.

As a research-first organization, the Institute will work on solving the most important and difficult challenges facing the creation of advanced robots. The Institute’s culture is designed to combine the best features of university research labs with those of corporate development labs while working in four core technical areas: cognitive AI, athletic AI, organic hardware design, as well as ethics and policy.

To achieve such advances, the Institute will invest resources across the technical areas of cognitive AI, athletic AI, and organic hardware design, with each discipline contributing to progress in advanced machine capabilities. In addition to developing technology with its own staff, the Institute plans to partner with universities and corporate research labs.

Not a lot of specifics here, which is how it usually goes with announcements like these, but we’re hoping for more details from BDAII itself next week. Meantime, we can speculate a little bit based on the context. Boston Dynamics as a company has (arguably by necessity) been shifting its focus for the last half decade or so toward robots that can reliably do things that are practical and useful enough that people will pay for them. This is a superimportant goal, not just for Boston Dynamics but for anyone doing cutting-edge robotics work, but it’s also much different from the equally important goal of doing fundamental robotics research that will enable the next generation of robots (like Atlas, perhaps) that can even more reliably do things that are even more practical and useful. Robotics as a whole needs both of these things to happen.

There’s no other way for commercial potential to happen: You can’t have a practical robot without having an impractical robot first.

The academic research community is of course a huge part of how subsequent generations of robots come into existence. But, there’s frequently a gap between the kind of research that’s possible in academia and the kind of research that’s possible in a commercially focused industry setting. That gap comes from a couple of different places—academia prioritizes original research that can be funded through grants and published, while industry prioritizes applications that promise some sort of tangible financial return. There isn’t a lot of space in either of those contexts for serious exploration of the practical applications of highly experimental technologies, and outside of rare contexts like the DARPA Robotics Challenge or SubT, the resources aren’t available to do so either.

This is what made Boston Dynamics (and perhaps to a lesser extent, Willow Garage) special: a combination of vision, funding, and time to drive critical aspects of the robotics research forward without getting bogged down in questions of practicality. But that wasn’t sustainable in the long-term for Willow, and it certainly seems as though Boston Dynamics has been trying to pursue two different paths simultaneously—one commercial path revolving around robots like Spot and Stretch, and a completely separate research path focused on Atlas. By separating itself from the part of Boston Dynamics that wants to sell useful robots and make money, it sounds like the Boston Dynamics AI Institute will be free to refocus on that very forward-looking research path, without having to constantly commercially justify itself. And let’s be clear—these advanced and experimental robots do absolutely have commercial potential, in the sense that technologies derived from them will be tremendously valuable in the future. It’s not always clear how or when that will happen, but there’s no other way for it to happen: You can’t have a practical robot without having an impractical robot first.

The next step for BDAII is to start scaling up, which means hiring as many roboticists as they possibly can. If you’re a roboticist looking for something new, here’s Marc Raibert’s pitch:

We plan to hire the best talent we can find and support them with the best resources, including equipment, facilities, computing, and staff, all wrapped in a culture that thrives on technical fearlessness, fundamental thinking, cooperative teamwork, respect for each other, and technical fun. If you have mad skills in robotics, machine learning, computing, software or hardware engineering and want to make a difference in the future of AI and robotics, consider joining us on what I expect to be an important and gratifying journey.

BDAII also plans to partner with universities and corporate research labs, which hopefully means that much (or at least some?) of the work that they do would be shared with the wider robotics community. But however it happens, this is a huge investment in the future of robotics, and I can only say it again: This absolutely looks promising.