Putting the Boston in Boston Dynamics with Marc Raibert

Hey everybody my name is Bryant heater I Am techcrunches Hardware editor I am Very pleased to be joined once again It's actually been a few years by Mark Raybert uh since his foundation in the Early 90s Boston Dynamics has become one Of the world's foremost robotics Developers and researchers Mark the Founder stepped down from the CEO role In 2019 and currently serves as the Company's chairman last year he became The executive chairman of Boston Dynamics AI Institute a Hyundai Sponsored organization focused on Developing Cutting Edge Ai and Robotics Applications mark thank you so much for Joining us once again Good to see you Brian So we're here to talk about the Institute we're here to talk about Robotics we're here to talk about Boston I think it's at the top of our list so So let's kick off with that uh I I Realize The name Of the company limits your Geographic Options a little bit but was it clear From the earliest days that Bossa Dynamics was always going to stay in the Boston area Well you know I'm kind of a lifelong Boston person I grew up in New Jersey But my mother grew up here so I had a Connection to Boston you know from from

A very early age Uh and you know I love I love Boston uh I went to school here You know left for 10 years and then uh You know eventually was faculty at MIT Uh you know I was a professor there for 10 years before starting uh Boston Dynamics and uh You know when we started it was halftime Boston Dynamics halftime MIT so we Started just next door so it wasn't Really motivated by Boston being a hub Of Robotics it was just you know where I Was and where I was comfortable and so We started here uh we did open a California office we have a California Office now Boston Dynamics does not the Uh The Institute that was a result of an Acquisition Uh not really uh we it was after we were Out of Google we uh Oh yeah yes it was a result of our Acquisition of kinema which was uh in Palo Alto I think when we first got them And then we opened an office in Mountain View which is still there But Boston is home is the basic answer I actually didn't know you left for 10 Years what did you do in the meantime Uh three years at the jet Republican lab Back in the 1970s And then uh six years at Carnegie Mellon On the faculty in both the robotics Institute and the computer science

Department Well well the second part makes a lot of Sense but were there were there a lot of Robotic applications for JPL at the time Uh it was fledgling but they had a a Mock-up of a Mars Rover this back in the 70s now it looked like a car and it had Uh I don't know if you remember the old Stanford arms the the ones that had a Sliding joint so it had one of those on Uh and it had some cameras and then There were a couple of different groups Working on it but in addition to that Tony bakesy who uh was a a researcher in Tele robotics he had a lab so I split my Time between Tony's lab and uh and that Mars Rover Group this was my first job After grad school So you did make that a move you did make That move to the other side of the Country initially Um and I've heard a lot of people talk About the the brain drain that cities Like Pittsburgh or Boston have have Undergone how large of a problem was That Exodus to places like San Francisco Or New York in the early days You mean of of our of talent yeah I Don't I don't think we've lost too many People to uh to the West Coast but you Know sometimes that happens when we were Part of Google there was a group of us From Boston Dynamics who moved out to The West Coast and I don't think I think

Those people mostly stayed uh and when We got out of Google they didn't come Back uh with us uh but you know Boston Has its own uh charms and attraction and There's lots of tech here lots of Schools uh doing good stuff you know There's MIT Northeastern has an Incredible robotics activity going when You come in April you should try and Visit over there it's very impressive uh Worcester Polytech has a group there's Holly Yenko at um Uh I'm I'm blocking on the university University of Massachusetts uh so There's a lot going on Yeah The waste lab yeah right yeah So when you have these conversations and And obviously you know with the Institute of being in play now with I Assume you're looking at a lot of you Know recent or soon to be grads from Schools like MIT what's your pitch for Roboticists and why they should stay in Boston Uh you know mostly we just focus on what The institute's doing and the work we Don't I don't think it comes up Explicitly as uh as a geography thing Mostly it's about what's the work going To be uh how committed are we to Research as opposed to uh product Development and you know that's the

Lever or the decision that most people Make do they want to be working on a Company that's developing a product and You know there's some advantages of that Uh for some people and then our long Suit is that we're working on you know Five to eight to ten year Horizon where We can work on uh problems where you can Really dig in on the technical issues And uh and do that and a lot of people That love that that idea Yeah So how dramatically has Boston and the Boston ecosystem changed in the 30 plus Years since the company was founded Well you know these days I don't even Know how many startups there are in the Area you know there's all the logistics Companies there must be a dozen of them Uh which all have uh robotics Talent uh At them uh There's uh the social robot activity you Know Cynthia Brazil's uh uh company and Uh and others like that and uh You know new things are springing up all The time the pizza company I can't Remember exactly the name that's that Was here Uh But you know to be honest I think more Globally I'm always thinking of us being Uh an international company and you know Now that we have we meaning Boston Dynamics have spot robots uh all around

The world there's about a thousand of Them out there now uh you know it's not Just all about Boston Sure yeah but obviously there are Certain benefits in terms of Infrastructure when it comes to running An Institute full of uh roboticists and AI researchers right well you know the Recruiting here is great there's Amazon That's got robotics people Google is Right across the street they don't Exactly have robotics in Boston but Obviously they have a lot of uh talent And Ai and software uh so those are Potential uh employees and you know There's a lot of technical people but we Recruit all across the country and and Somewhat from Europe uh so we've been Getting uh you know applicants from all Over Yeah obviously the conversation is Different when you're trying to track Somebody to your company versus a Student at a place like MIT or Harvard That's looking to found their their own Startup Um you know insofar as you're aware of It that as a former faculty member Places like MIT and CMU our schools like That doing enough to help students found Startups Well you know when I started Boston Dynamics one of the key events Was uh the MIT hired a technology

License office office person uh who had Been at Stanford Stanford at that time This was 30 years 30 something years ago Uh was ahead on the curve of stimulating Spin outs and uh startups and this guy Moved they hired him at MIT and I can Remember going into a meeting where I Was totally skeptical that my future Could possibly involve leaving MIT and Starting a company and in a one-hour Meeting he uh you know he changed my Mind mostly because he said Um If the founding people if the people who Did the invention aren't bought in and Excited about the the company it doesn't Have a chance of success and so the Whole program that they had at MIT was Designed to motivate and Inspire and Really support uh the founders uh and When I I just expected that MIT would be You know hugging the IP like this and Instead they were saying you know Take it as far as you can and I think That's different than some schools you Know we've negotiated we being a company Have tried to get IP from from some Universities and they're really holding It close with very restrictive terms and All that and uh you know it's Discouraging when you get in that Situation but anyway so uh so that was Uh I can't remember your question but The one thing that happened was MIT

Encouraged that Yeah yeah it was uh you know explicitly About sort of like incubating and Accelerating startups but it's it's Interesting to hear that they were it Sounds like they were also ahead of the Curve by a couple of decades Yeah yeah and that that uh that really Made some things happen there were also Other activities like the the 50k Activity at MIT where someone had uh Every year they would uh fund startups Uh based on their evaluating their Business plans oh the other thing I was Going to say is there's Olin College uh I don't know if you know Olin but it's Another area College it's a very Interesting College because uh Everybody's tuition is paid for by the School Uh so that's kind of an uh an equalizing Thing but they have a very strong uh Development of entrepreneurial uh spirit And uh We've you know we've had a number Of interns at Boston Dynamics from there And they were always gunning to you know To start their own company that was Their uh their plan so I think it's been A a force here in the Boston area as Well So obviously your your wardrobe hasn't Changed since you left the CEO a role at Boston Dynamics but uh how is your Day-to-day changed since leaving

Well you know when there was a point at Boston Dynamics where we were clearly Turning more commercial uh and I decided That I wasn't the right guy to to do That I'd also heard it talk uh by um Fortunately I can't remember his first Name his last name is uh blankenfine who Was the chairman of uh is it Lance JP Morgan or Goldman Sachs one of those Places and he gave a talk about his Decision to uh to step down and he said You know you really need to step down When things are going great if you step Down when things are in trouble then They think you're a bum and they kick You out but if you do it when it's going Great it's hard because things are so Much fun uh but that's when you have to Do it so I listened to that and decided He was right and that plus the fact that We were turning to commercialization so I became the chairman and uh appointed Uh you know Rob plater had been my right Hand man he'd been my graduate student At MIT been with me for 30 years almost 30 years at that point 27. so he became The CEO And uh And you know I I didn't have enough to Do to be honest and then covid came so You know everybody is even withdrawn Further and so it was kind of sleepy and Slow and you know I thought about Retiring I'm old enough to retire

Uh and they're certainly dressed for Retirement Well I've always been dressed every time Um I mean that's one of the things about This life it's it's um It's fun and work at the same time you Know what could be better than that uh But then uh when uh Hyundai agreed to Fund The Institute I decided I needed to Be here every day to work and to inspire The others and it's been great it's it's Just like being back in it full full Bore so uh uh I've been loving it been Working hard uh And uh I couldn't be happier How much of that pushed towards Commercialization or push towards Productization was a result of the Hyundai acquisition Um it really started I think the tip of It came when we were still at Google Uh but then SoftBank you know the first We were with SoftBank for four years I Think and the first two years it was Just all about the future and then uh You know it got to be more about uh what The products were and uh you know Tightening up the the bottom line and That kind of stuff and when when we got There uh you know we turned but I think In other force is that Rob plater has Always wanted to uh commercialize I I'm A research guy uh I really want to work On the long term and you know make the

The Next Generation or the generation After that happen of robots uh but Rob Is really uh you know interested and Committed to the commercialization side I think you might have seen I think There was a Boston Globe article that Said the company uh was all about making Money now uh And uh so I think it's been a great uh You know a great path where I'm going on The path I like and the the company is Continuing you know to do do great work Including research uh as you can all see Yeah I I mean obviously there's it's Good to have a company that is Self-sustaining and that can generate Its own money Yeah So was the AI Institute was that part of The Hyundai deal from the beginning No Um so the origin of the Institute is uh After I became chairman and started to Get bored I wrote a proposal that I Actually shopped around to a few Billionaires and uh and got someone to Uh to agree at a level that wasn't Enough in my opinion uh then covet Happened so I couldn't you know go and And visit people as easily so I kind of Slowed down but then after the Hyundai Deal happened it was all done I I Pitched it internally and uh and they Went for it and uh you know and here we

Are So beyond you know the obvious Contribution of uh 400 million from the Start what is Hyundai's relationship to The Institute Well they are you know they're the only Stockholder uh at the in the current Arrangement Um you know it's been very collaborative Uh I talked to there's some managers Involved but I also talked to the Chairman on a regular basis of Hyundai He's a real Enthusiast he's he's really Uh you know leaning into the future Where uh software uh Ai and other High-tech things EVS are you know Central to to Hyundai to HMG and uh you Know we're just a part of that leaning In process both Boston Dynamics and uh The AI Institute It is pure research is that a hard thing To sell to a you know a giant Multinational automaker Uh you know it's only a year it's Actually the institute's only been Around since last summer although the It's been in the works for over a year Uh You know you never know until later uh What the staying power is right now my Pitch is to avoid products because Products force you into quarterly and Annual work products force you into All the various needs of all the various

Customers And you know they have a lot of good Information but they also take you in Lots of directions And if you want to you know do the Vision of what's going to come next I think it has to come from the Technical people uh who are developing It uh and so uh The Institute right we You know I proudly say we're not doing Products you know if Downstream does Hyundai get bored with that I don't know but this is what we're Doing And basically no one's trying to Um To steer me at the at the moment This this philosophy of of avoiding Products was that was that part of Boston Dynamics mission statement at the Beginning Boston Dynamics no uh okay no and you Know we had products Um we had some software products in the Very early days that uh that we actually Made money on uh you know in the early Days we were always in the black because We didn't have any investors up until Google acquired US 20 20 something years In so we always had to be making money Some of it was a lot of it was contract Work but we also had some software Products not robot products but software Products

Um so no we didn't we didn't say we're Not going to do products so we're not Going to make money uh in those days So aside from sharing a name what is the Institute's relationship with Boston Dynamics Uh there's dude I mean we're just Friendly but um we don't have any IP From Boston Dynamics it's really just uh Sharing the name uh at the moment and of Course I mean well and I'm on the board There Um we haven't we've been we haven't been Recruiting from there although we have a Few people uh most of whom uh uh either Left on their own or something like that And of course Al Rizzi who's the long Time uh Chief scientist there is the CTO At The Institute uh he kind of we we Came together uh to it so we're just Loosely connected So if product if productization and you Know commercialization aren't focuses For for The Institute What happens with the IP and the patents And the other things that are developed In The Institute I mean we have a a multi-prong plan uh We can do spin outs Which it's interesting spin outs are by Some are seen as a way to commercialize For me it's a way to protect the Institute from productization from Products just to sort of push them out

And let them do their own thing away From the that's right because you know We have really smart people who are Inventing stuff and you know you know That they're going to be a product that Good product ideas and so we're we're You know we expect to spin them out uh Maybe you know maybe incubate them a Little bit if that's what they need Um There's also licensing to possibly the Boston Dynamics possibly to uh to Hyundai or third parties you know we're Nowhere near that yet uh you know we're Only a few months in so we don't have Any any real IP at this point we just Have talent at this point sure but you Know I assume these are conversations That that happen early on but it sounds Like there's no foregone conclusion that Anything that's developed by The Institute is going to end up at Boston Dynamics or Hyundai no No that's right So you know we discussed the difference Between the the work being done at you Know Boston Amex which as you said You know is doing research but now is Also very much focused on bringing Products to Market but how does the Institute's research differ from the Work at universities like MIT or CMU Well first of all let me talk about the Structure

Um let's see first let me talk about the Difference with awesome Dynamics I think There's two main uh differences one is We're aimed five to eight to even 10 Years out and I think most of the work At Boston Dynamics is aimed at you know Uh what's the next product well Satisfying the current set of customers What's the next product going to be and Uh And you know developing on a shorter Time frame uh as you need to if you're Commercializing and uh and making Product The other difference is Boston Dynamics Is pretty squarely focused on the Physicality of the robot including the Ability to create behavior that responds With its environment But less so on the AI side of things how Do you communicate with the robot how do You make robots that can plan things uh You know basically robot all robots Today are about as dumb as doorknobs and People see a robot doing parkour and Backflips and they think that's just Like a person Uh but in almost every other dimension It's not anything like a person you Can't communicate with it it can't write Its own programs uh Its perception is usually limited to the Things right around it in its Environment and one of the things one of

The things I appreciate that Boston Dynamics has done that I feel like more Companies should do is release the Longer cut you know release the bloopers Release like all of the all the Difficulties in the trail and air that Went into making those beautiful parkour Videos Well you know back when I was just a an Early professor in the early 80s I made A robot that balanced itself you know a One-legged pogo stick robot and I showed That And no one cared and then I showed one Where I showed it falling over how Quickly and easily it tips over and then All of a sudden everybody appreciated oh Uh that was intro that was hard that was Interesting so the you know the bloopers Are a great tool for making people see Uh the accomplishment in a in a bigger Light it also shows the frailties but it It also strengthens the uh the Accomplishments so I agree So you know again we discussed a little Bit of the differences between Boston Dynamics and Company and Boston Dynamics The Institute but obviously there's so Much great work being done in Universities right now what what Opportunities are you afforded at The Institute that they don't necessarily Have at those schools Well I think in The Institute you know

You have a university which is full of Smart people with uh Blue Sky ambitious Skulls to do uh new novel things And then you have a corporate lab which Has uh teamwork they have uh support From Hardware software sensor electronic Engineering uh And they have uh schedule and a budget Discipline And I think of the Institute as being Right in the middle having larger scale Uh teamwork we plan to have about 50 People in engineering that supports the Work of the researchers in addition to Other engineering people in the research Groups uh while we're trying to maintain That you know future how to make the Future happen and solve really the Fundamental problems in robotics not Just the feature that the that is needed Right now but how do you really make Robots smarter so they understand the World they're operating in so you can Say you know see that guy over there What he's doing uh do that instead of Having uh a dozen programmers work for a Month to get that programmed So that's what we're trying to do and I Think the scale of having uh those teams Is really going to be important so you Asked what that's going to let us uh do I think we'll be able to you know Obviously uh chat gbt is all the rage And I don't think chat TBT is the right

Thing for For many of the tasks that robots do but The idea of a foundation model tuned to The functionality of what robots they Need to do in terms of their perception Communication planning I think that's a Real possibility but to do that you have To work it at a larger scale than most University Research Labs can do also to Build Hardware you know I think part of Boston Dynamics is success is that They've built invested in the team that Can build really high performance and Reliable Hardware uh we're gonna we're Gonna do that too because I think you Need to have the software and the Hardware uh working like this in order To succeed in this business Foreign The concept of uh multi-purpose humanoid Robot comes around every few years Uh you've been involved in one of the More notable examples uh the atlas robot Does that that bipedal form factor Ultimately make a lot of sense Uh you know I don't know uh Elon thinks So that's that's interesting Um I would rather point at spot as a a General purpose robot just because you Know we designed it without a particular Application in mind it's a platform Where you can customize it for your use You know we have a thousand of them out There which are being used in a

Reasonably diverse set of uh Environments and so you know we're Finding out how well uh that works uh You know can you amortize the the Technical investment in the platform and Have it really pay off for a a wide Variety of uh of use cases Um you know at the Institute we're in a Planning phase where we have a list of Projects uh that we're contemplating Doing and we're getting all our talented People to say things about what they'd Like to do and getting the balance Between trying to have robots that do Everything and robots that do one thing Is you know a hot topic for us and we're Probably going to pick a couple of Activities that span that space and you Know see how it works out Yeah certainly there are a lot of there Are a lot of stops on the path to a you Know a real general purpose robot but I But I have to ask you know you brought It up What were your thoughts on the Tesla Optimus demo Oh uh I thought that um Uh they they'd gotten a lot more done Than I expected Uh and they still have a long way to go Uh I thought that their design was uh Was it was interesting that cosmetic you Know the second robot that wasn't really

Working but was on stage I mean Cosmetically it was very interesting and Uh Uh you know I'm a I'm a Tesla driver I've had a couple of them uh I really Admire uh Elon uh you know despite the Recent Twitter activity I think uh you Know he's a brilliant guy and uh I Absolutely wouldn't count him out And we've only got time for we've only Got about a minute left and I have to Ask you why was it important that the Company signed the weaponized robotics Pledge You know I think that there was a lot of Sentiment among the employees that uh That we should um Uh you know be uh be good about uh the Robots you know the robots aren't Suitable for weaponizing uh there's so Many additional concerns if a robot were To be weaponized and uh we just wanted To say uh you know we shouldn't be doing It with ease I think there was some Unhappiness that it looked like uh some Of the other companies had uh just kind Of flung a weapon on there without maybe Necessarily considering all the safety Things I mean as much as anything it's a A friendly fire type worry you know if You have something that's that's not Doing what it's supposed to do Great uh unfortunately we are all out of Time

Um Mark I appreciate it as always and Hopefully I will see you in a few weeks In Boston great good talking to you Take it easy