Michael Li grew up in 1960s China and studied computer science at Tsinghua despite never seeing a computer. He moved to University of Waterloo for free applications and worked on algorithms for Canada's robotic arm on the International Space Station. He watched AI engineer salaries explode from scarcity to over $2 million.
Chapters
Well, hey, good morning, everyone.
Good morning.
It's very good to see so many young, energetic faces here. I really like your setup here. I want to sit there. We should be sitting there.
So Michael, before we start, I'd also like to just get us— since I get to know you, even though I've known you for many years, so they get to know you. So maybe you tell us a little bit about Your background and something, a fun fact about you. It could be your favorite food or a hobby that no one knows about or a special talent. So with that, background first.
Where were you born?
Where did you go to school?
I don't know how much useful knowledge you can draw from my experience because I'm certainly the oldest guy here. And, you know, the—
Not by very far.
But seriously, the environment I grew up with is really far from what you guys grew up from. I was born in northeast China in 1966, so you figure my age already. I look— certainly look younger than I am.
That's true.
I presume.
One—
there's a couple of things that I'd like to mention. When I applied for undergraduate study at Tsinghua University at computer science, I didn't even know what a real computer looked like.
But you applied.
I applied simply because of the reason that I read a few articles about how to program in BASIC language, and I feel like this is something interesting. something that I like to study. In, you know, the common conventional knowledge about what to study back then in China, you know, all the best students would study engineering or, you know, science. This is a pure stereotype in China back then. So you either study math or physics or, you know, chemistry. Computer science was totally new back then in 1984. So if you guys figure 1984 was the year that IBM released its first personal PC, before that it was only iPod 2 available on the market as a popular phone. So that's, that's, that's, you know, how I started in computer science.
So do you feel like education back then was— it's really prepare someone for employment?
Not necessarily—
or is it—
We didn't even know. Let's just speak— not much job related to computer science. No, really no. So you didn't even know this would be a good or a more practical— It was nothing like that. It was not practical consideration for me to get into computer science. It was a lot of information disparity. It was purely out of my interest. And this is something I feel that I like to do. And I didn't even think of how I should be employed. And the other thing you guys probably don't know, back then in China, college students are assigned to jobs. You don't find your job. The country, the government will assign you jobs. So I didn't need to worry about whether I would have a job or not. It's a planned economy. Yes, it's a planned economy. So I'm trying to say is, back then, you don't get help from your parents because they don't know anything about engineering. And you have very little knowledge about what you're trying to study. But I didn't regret the fact that I got into engineering school because I think I was purely driven by interest. And that's something truly important if I look back. If any lesson that I can draw, I think that's— That's a good lesson. Probably same, I like to say.
Mm-hm.
So the other interesting thing I like to say is that after graduation, I was interested in studying abroad, but based on a particular political event in 1989, you guys probably can figure it out. So I was kind of desperate. I didn't see my future in China back then. So I think maybe I should explore studying abroad. And I try to find a— the best book I can find is a college guide. I forgot the name. It's popular.
Is it US News?
Yeah, it's US News, something like that. US News something.
Yeah.
So the first thing I do is I look at all the colleges. No, not rankings. The only criteria I can look at is I can only apply for the college that does not require administration fees. Usually, typical admission fees was like $50, $80 back then.
How much was the average wage in China at the time?
It was about $20 per month.
So it would be equivalent to a—
3 months' salary for college.
3 months' salary for a person.
Yeah, so that was purely, purely my idea. In Silicon Valley, that would be what, $30,000? Yeah, $30,000 back then, yeah, $40,000.
So imagine applying for one school, it's $30,000.
Yes.
You wouldn't apply to too many schools.
I wouldn't even apply for, you know, 2 schools. I actually borrowed money to take my TOEFL test. Otherwise, I wouldn't have even qualified for application.
So these guys watched the movie American Dream in China. It's about the Xindongfang.
Yeah.
You know, the English prepping school. Did you go to one of those?
I didn't.
Because it didn't exist then.
Correct, it didn't even exist. It was '89 when— it was 1990 when I was trying to apply. So I ended up in University of Waterloo. I was very grateful because—
Waterloo?
Yes. First of all, of course, it didn't ask for admission fees. So that's why it's on my shortlist. And the second of all, I think my professor was a very influential professor in the area of machine vision and artificial intelligence. So I was very fortunate.
This was before it was very popular.
It was before it was very popular. Well, you know, I studied artificial intelligence at college. So it's been up and down for a couple of times. This is now the first time that AI becomes very popular.
The going rate of AI engineers in Silicon Valley, do you know how much they cost?
Last time I heard, it's about $2 million plus.
Wow, you're very well kept up to date.
Yeah, I'm in the industry that information is really important. That's why I'm trying to say is information was a luxury back then in China, but still you need to manage to get most out of the limited information you get. And that's why I think I was very fortunate to end up where I am nowadays.
And the investment in the education in Waterloo was a good decision?
It was a good decision based on my limited resources and information.
Yeah, zero admissions fee. Applications fee, and you could be making $2 million a year now as an AI engineer.
That's— yeah, don't make me regret that. I probably made the wrong decision. Yeah, that's— I didn't even know that it's going to make so much. I was a little concerned about my employment prospects back then. You know, AI was not particularly popular when I graduated from— got my master's degree.
It was probably very academic, right?
It was very academic. There wasn't many industrial applications back then. I was involved in— at school, I was involved in an aerospace project. You guys know the ISS, International Space Station, right? Canada was one of the 6 nations that got involved. Canada was responsible for doing the robot arm. You guys seen the picture, right? There's a big Canada sign there.
Mm-hmm.
And our lab was responsible for delivering the vision system for that. So there's a couple cameras. So I was the guy writing all the algorithms behind the camera, trying to capture and position the objects in arbitrary positions so quickly you can grab that. So that was my contribution to that.
Did you see the movie Martian?
Yeah.
How much of that is true? Bring back some memories? They spent a lot of— there's, what, 10,000 people working on supporting this one guy? Yeah.
Well, I was a very small part of that. I didn't have much big picture views, but I know it's a huge project. The country spent hundreds of millions on that project. Just for the arm. That's a very big project. I believe that my algorithm is still being run on the space station.
So I'm pretty— I'm sure.
I'm sure. Yeah, I'm very proud of that. They don't change that often.
Did you embed some funny comments in there?
Yeah, I would not tell you. There's some backdoor.
