In the Media

Dr. Eliasmith on Tesla's Optimus Robot: Good Morning Hamilton

March 30, 2023
Dr. Travis DeWolf

Throwback to Dr. Chris Eliasmith's insightful interview on Good Morning Hamilton in 2022, where he shared his thoughts on the latest developments in AI, robotics, and Tesla's Optimus reveal.

Rick Zamperin: This is Good morning Hamilton. You're listening to 900 CHML, Rick Zamperin with you, waking you up on another beautiful morning in our city. Did you get a chance to check out Optimus? Not Optimus Prime. Not the transformer. No, this is Tesla's new entry into the humanoid robot world, so to speak. Last Friday, Tesla unveiled this robot it calls Optimus and technology insiders, at least all the ones that I heard and saw from, are giving it a big thumbs down. Why? Chris Eliasmith, he is the director of the Center for Theoretical Neuroscience at the University of Waterloo and the Canada Research Chair in theoretical neuroscience. He joins us now on Good Morning Hamilton. Chris. Good morning. How are you?

Chris Eliasmith: Good morning, Rick. I'm doing great.

Rick Zamperin: It's safe to say that Optimus fell flat.

Chris Eliasmith: Not quite literally, but almost.

Rick Zamperin: Yes, it almost did. One criticism is the robot has human-like hands with five fingers. You know, one AI insider wasn't too thrilled about that. Another one called it cringe-worthy. What was your first impression of Optimus?

Chris Eliasmith: Yeah, underwhelming really is the keyword of the day. When you see it, walk and move, it looks like something from about 10 years ago. So you know, the impressive part is how quickly they did it. But what they ended up with is not impressive.

Rick Zamperin: Given that, are you surprised that Tesla rolled this thing out?

Chris Eliasmith: Not really. I think some people have suggested that it's really much more of a recruiting effort than anything else. I still think it's very strange. I understand where the question is coming from because, you know, when Tesla put out their cars, they were super special because they were electric. Or when Elon Musk launched SpaceX, it was super special because it can land and be reusable. But there's literally nothing special about this robot.

Rick Zamperin: Aside from the recruiting efforts, what do you think Tesla's end goal is here? Is this to create a personal butler or maybe a robot that is used by businesses or governments? What do you think?

Chris Eliasmith: Yeah, I think they have a pretty practical sort of goal in mind. And that's to replace the people that are in their warehouses and in their factories with robots that, you know, don't ask for more money and don't complain about human rights violations and so on.

Rick Zamperin: Yeah, and steals jobs from living, breathing souls. Any guesstimate on how much this perhaps might cost when it is officially released?

Chris Eliasmith: So they are suggesting it will be $20,000 per robot. I would be extremely surprised if they could get something that cheap. One of the sort of cheaper, actually functional robots on the market right now is about $75,000. And it's much smaller with fewer parts and so on. So they've suggested it will be 20k. But I'm guessing it'll be four to five times more than that.

Rick Zamperin: Aside from the factory floor, where else are robots these days?

Chris Eliasmith: Well, lots of people have them in their houses, you know, I think the little Roombas that are zipping around all over the place, they count. We don't see them in too many other practical scenarios. In fact, we can sort of think of self-driving cars as getting there. But you know, we obviously don't have fully autonomous vehicles yet. And yeah, for the most part, you really do see them still confined largely to factories and so on. I guess the other big place is, of course, things like space exploration and very dangerous on Earth missions. When there are things like nuclear meltdowns, you're much more likely to see robots in those sorts of scenarios.

Rick Zamperin: Yeah, those bomb detonating devices that police will use, obviously, would be in that category as well. Good morning Hamilton and our guest is Chris Eliasmith. He's the director of the Center for Theoretical Neuroscience at the University of Waterloo in the Canada Research Chair in theoretical neuroscience. Humanity, as you know, has had a decades-long infatuation with robots. How do you think that's evolved over the years?

Chris Eliasmith: That's an excellent question. Initially, the very first robots that showed up in film and so on were dangerous creatures. And that has really stuck around, which is a little bit surprising when you see things like Optimus show up that can barely walk. But, you know, of course, the Terminator carried that idea on and so on. And I think we're just sort of evolving to the point where we realize that it's going to be very difficult to make something as dangerous as what we initially imagined they would be. But nevertheless, people are being very careful about that. There are many sorts of letters which scientists signed to say, you know, “I'm not going to allow any of the technology that I developed to be used for military purposes,” and so on. But I think the really practical robots we have right now are anything but threatening, but also not quite as useful as we hoped.

Rick Zamperin: So Optimus is not bringing us one step closer to robotic world domination?

Chris Eliasmith: Not really. It can barely bring itself one step closer to the edge of the stage, from what I saw. 

Rick Zamperin: Yeah, I think it was three Tesla employees carrying a next-generation Optimus, I guess a superior Optimus, because it couldn't walk.

Chris Eliasmith: Exactly, yes. So the first one they showed was just, you know, kind of a metal husk that was what they called a research platform. And then the advanced one was really like on a pogo stick that they could roll around the stage, and it just kind of swung its arms and legs around. So presumably, there are advances there, but like I said, we have no idea what they are. There doesn't seem to be anything special, unlike Elon's other companies.

Rick Zamperin: Very interesting stuff. Chris, thanks for your insight into this.

Chris Eliasmith: My pleasure.

Rick Zamperin: Chris Eliasmith is the director of the Center for Theoretical Neuroscience at the University of Waterloo, and the Canada Research Chair in theoretical neuroscience. You got to Google Tesla Optimus robot. There's also a story on Global, in which it's featured. To me, it's in the embarrassing category. I mean, here is a company – here is a multi-billionaire who is launching this product or this prototype, and it's not even, by robotic standards, fully functional. I don't know whether they're putting the cart before the horse or just wanting to get this thing out there to say, "Hey, look at the new shiny toy we're making," or whatnot, but wow, I got some work to do, and it's for sure.

Dr. Travis DeWolf is ABR's Sr Research Scientist II and Co-Founder. He has a PhD in systems design engineering with a focus in computational neuroscience from the University of Waterloo. He is one of the co-founders of ABR, and is an expert in the design of neuromorphic and edge AI applications and building computational models of the motor control system in the brain.

Similar content from the ABR blog