Apr 16, 2024 · 47 min · 14 segments
Are you ready to go behind the scenes and uncover the secrets of the company playing a pivotal role in the future of autonomous and electric vehicles? In this riveting episode, we dive deep into the…
Grayson BrulteHost
I recently had Alex Kendall, the CEO, co-founder of Wave on the podcast, and he told this great story.

Do you feel that AV 2.0 approach is the future of autonomous driving for the next five years, perhaps 10 years?
I have a Bill Gates story as well, but I'll save it until after Peter answers the
So I think in general, right, let's talk technicals for a second, and then I'll answer the specific question.
So the field of machine learning is all about function estimation, right? Fundamentally, what you're trying to do is you're trying to, instead of using, let's say, traditional math and handwritten code to create a function, you're trying to effectively fit a model to produce the correct answer that you want.
And then the question is, well, how big can that function be? The earliest applications of machine learning were very small functions solving very small, distinct problems.
And the trend in ML has been, well, we can keep making this function estimator larger and larger and larger to estimate more and more complicated functions.
That's enabled through not only a number of breakthroughs on the research and the science component, but it's also enabled through breakthroughs in chip technology and the fact that you can fit so many transistors on multiple processors.
And so the question of, well, how big of a function can you estimate? I think the answer is going to be, well, it's probably going to keep growing.
Every year, we're probably going to be able to estimate larger and larger functions.
And then the question is, well, what does it take to productionize these large function estimations? And that's really where our expertise is incredibly deep, which is making these systems safe and and really enabling lots of companies to be able to do this and bring these systems to production.
And so I think there is an inevitability, just like Moore's law, there will be more transistors available on these processors that's going to keep getting larger and larger, especially with the trend towards chiplets and being able to now combine actually separately manufactured silicon together on a single package.
The limit to how big these chips can be is going to keep getting larger and larger and larger.
And my light Bill Gates story is a couple of years ago, Eric Tornberg from Village Global set up this one-on-one with Bill.

I recently had Alex Kendall, the CEO, co-founder of Wave on the podcast, and he told this great story.

Do you feel that AV 2.0 approach is the future of autonomous driving for the next five years, perhaps 10 years?
I have a Bill Gates story as well, but I'll save it until after Peter answers the
So I think in general, right, let's talk technicals for a second, and then I'll answer the specific question.
So the field of machine learning is all about function estimation, right? Fundamentally, what you're trying to do is you're trying to, instead of using, let's say, traditional math and handwritten code to create a function, you're trying to effectively fit a model to produce the correct answer that you want.
And then the question is, well, how big can that function be? The earliest applications of machine learning were very small functions solving very small, distinct problems.
And the trend in ML has been, well, we can keep making this function estimator larger and larger and larger to estimate more and more complicated functions.
That's enabled through not only a number of breakthroughs on the research and the science component, but it's also enabled through breakthroughs in chip technology and the fact that you can fit so many transistors on multiple processors.
And so the question of, well, how big of a function can you estimate? I think the answer is going to be, well, it's probably going to keep growing.
Every year, we're probably going to be able to estimate larger and larger functions.
And then the question is, well, what does it take to productionize these large function estimations? And that's really where our expertise is incredibly deep, which is making these systems safe and and really enabling lots of companies to be able to do this and bring these systems to production.
And so I think there is an inevitability, just like Moore's law, there will be more transistors available on these processors that's going to keep getting larger and larger, especially with the trend towards chiplets and being able to now combine actually separately manufactured silicon together on a single package.
The limit to how big these chips can be is going to keep getting larger and larger and larger.
And my light Bill Gates story is a couple of years ago, Eric Tornberg from Village Global set up this one-on-one with Bill.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.