Mark Blyth
speaker
294 appearances
3 recordings
3 series
first heard Aug 2024
last heard 3 Jul
Mark Blyth’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.
Appearances
Yep. Seriously. Now, in some places, it may take about three years. TechSoup probably has easier permitting. But the point is, if you're not making this stuff now, you don't just turn on a tap overnight and make it.
It's wishful thinking. If we change this price, magically everything will appear. No, I don't think so. I think you just end up with a bunch of shortages and inflation.
And that's what Biden is trying to do with the Inflation Reduction Act. And if you think about it, Biden didn't take down any of Trump's original tariffs. Trump mentioned that in the debate. And why not? Well, because ultimately, if you're going to rebuild an industry that China's got leaps and bounds ahead in, you have to have tariffs. Otherwise, yours will never be competitive.
The real problem with us is just basically, though, they're so far ahead and the technologies are so well known that things like with solar panels, I mean, we could spend billions making American solar panels. And they'll never be as cheap as the Chinese stuff, because ultimately, it's a pretty simple technology.
And they have those huge firms, state-owned enterprises, with all the critical minerals and the polysilicon. It's all made there, right? To make our solar panels here, we need to import the polysilicon from there and turn it into a panel. Just let them do the panel. It's so much easier for them to do the panel.
We should be trying to do the stuff that's higher up the value chain, carbon capture and storage, small nuclear reactors, all the stuff they can't do. But that doesn't supply as many jobs, and it's a bit more of a wild bet. Right.
I would say, did you really evade bad inflation? Was the raise from 0.5% that instrumental in causing the fall in inflation? Or was it kind of like performing a circus trick at the side of the circle while the real action is going on elsewhere? In other words, it was supply shocks. They come along. They dissipate. You're fine. Were the central banks really in control of this?
Were the ones whose magic wand and Jedi mind tricks over expectations really what was driving the show? Maybe it was. I can be open to that argument. But there's also a bunch of other stuff that's going on. So the biggest problem you ever have with diagnosis is a bad diagnosis. Any treatment with a bad diagnosis is not going to work. I think we have incomplete diagnosis.
I think that we need to be more thorough in our diagnosis of what's going on. Because if at the end of the day, you really believe the story is like there's one tool, one hammer, one nail, interest rates recession, kick it out of the system, what you're saying is some of the most vulnerable people in our society are the ones that are always going to have to pay for the costs.
The control valve becomes unemployment. And it's not my unemployment. And it's not your unemployment. It's their unemployment. Right. That is the stakes of them being wrong if, in fact, they're wrong. Exactly.
Well, when you spend the past year writing a book on it, you're like, do you think you know one or two things about that? If I don't, I'm in trouble.
This is 20VC with me, Harry Stebbings, and we're sitting down today with one of my favorite writers in AI. He's been a big proponent in the belief that despite what many people think, increasing the amount of compute from this point will be unlikely to increase model performance significantly moving forward.
I'm thrilled to welcome Arvind Narayanan, Professor of Computer Science at Princeton and the Director of the Center for Information Technology Policy. This is an incredible discussion that goes very deep on the bottlenecks in AI today, and you can watch it on YouTube by searching for 20VC.
You have now arrived at your destination. Arvind, I am so excited for this, dude. I was telling you just now, I am one of your biggest fans on the Substack newsletter. I can't wait for the book. So thank you so much for joining me today. Thank you. I really appreciate that.
Now, I want to get pretty much straight into it, but for those that don't read the substat, which they should do, can you just provide a 60-second intro, some context on why you're so well-versed to speak on the topics that we are today?
I would just love to start because you have done some work in the cryptocurrency field, done a lot of work in the crypto field. I'd just love to start before we dive in deep on infrastructure. How does the AI hype today compare to Bitcoin hype? How is it the same and how is it different?
I do want to ask, and we'll start with kind of the hardest question of all, but it's the most important, and you've written about this, and I loved your piece.
The kind of core question that everyone's asking right now is, does more compute equal an increased level of performance, or have we reached a point where it is misaligned and more compute will not create that significant spike in performance? Kevin Scott at Microsoft says, absolutely, we have a lot more room to run.
Why are you skeptical, and have we gotten to a stage of diminishing returns on compute?
MARK BLYTH, Can we just take them one by one there? There was a lot of great things that I just want to unpack. You said there about kind of potentially the shortage of data being the bottleneck to performance.
Showing 221–240 of 294 · page 12 of 15
← Previous
Next →