YouSaid · the spoken record

Mark Lemley

lines on the record: 69
first: 2020-03-09
most recent: 2020-03-09
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“Exactly. There is law. It's a lot of it was made in the 1960s with a very different concept of space. almost all of it with the idea that everything up there is governmental. And as space becomes a kind of place we are going and we're increasingly going with private and not just public moves, I think there's going to be a lot of interesting rules from sort of staking claims to mineral rights and media rights to intersections and collisions and orbital paths among the increasing number of satellites that are up there and so forth.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Well, so it might be that it goes increasingly outside the court system, right? And that sort of doing this by having a bunch of jury trials doesn't make sense, that we ought to have kind of regulatory rules and standards. Those two can be problematic. My favorite example here are the self-driving long-haul trucks, which are out there on the roads already. They are carrying stuff. But they have to shut down periodically. Because the way we allowed them into our existing legal regulatory system is we defined the truck itself as the driver. But commercial truck drivers have to take rest breaks from time to time or they'll get tired. So we literally shut the truck down for eight hours every so often because the laws require us to let drivers take rest breaks and the driver here is the truck itself.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“And so when the car swerves left and hits a pedestrian and we don't know why, everybody's going to point the finger at everybody else. But maybe we ought to just sort of get rid of the question of who's to blame and focus on the outcome, right? And you know what? You hit a pedestrian that's bad. We should compensate. We should pay for that. And we should make sure that if your car hits too many pedestrians, there's something wrong with it. We need to shut it down. We need to change the algorithm in some way. But kind of trying to decide sort of whether you'd made the morally right decision, which is the way I think we think about it right now doesn't make a lot of sense for robots.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“They will own the robots, and so we can hold the companies liable. Although, boy, I'll tell you, that's going to be a can of worms because it turns out that the person who wrote the algorithm and the person who wrote the training data set and the person who decided what training to do and the person who owns the car are all going to be different people.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“There's a lot of our legal system that's actually really about that. I feel better because I have punished you. You have suffered for the bad thing you did to me. But it doesn't make any sense to talk about doing that with robots. And I think sort of we've designed a system that assumes humans and human behavior. And as we start to accommodate that system with humans.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah. And so I think one of the things that changes is we have a sort of tort system. What happens when my car runs into yours? We've got a system that's based on figuring out who's at fault. And that's based on state of mind. Was I taking enough care? Was I drunk? Was I exactly? That doesn't make a lot of sense when it comes to robots, right? And so I think we're going to have to rethink the way we think about accidents and maybe rethink our kind of whole notion of fault and punishment because punishment also doesn't make a lot of sense. Christina Mulligan has a paper in which she suggests maybe she should be able to punch a robot that injured you.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Right, in that kind of classification. All right, finish talking about the autonomous cars. There's all kinds of that with robots and things like that. We're not getting into robot rights, but this idea of what changes as we have autonomous cars or technologies like that in our society.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Well, that's the thing, right? And I think it's probably in the 1990s humans could outperform a computer at doing that. Probably today a computer well outperforms humans in that kind of classification.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“People do not realize. I would spend a time in those places with them. It was fascinating. It was a human decision based on where it would go. Of course, it was not scalable, nor was it correct most of the time.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, so let's say your credit score, right? I am denied alone. I am denied alone because the algorithm said, well, sorry, you're not credit worthy enough. I want a human being to look at my loan application and say whether or not I should get it. I'm not sure that that's actually better. In some ways it could be worse. Exactly. If I were, honestly, if I were a black man driving at night in some parts of the country, I think I'd rather have a computer decide whether to stop my car than a cop. There's at least the possibility that we can actually get around, avoid some of those biases. But people worry about it. People want to sort of...”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, so a lot of people are thinking about the problem of robotics and AI from a legal perspective. And there are a variety of pieces to that. I mean, everything from who's responsible when you're self-driving car runs somebody over to the problem of figuring out whether your algorithm is discriminating against people when we can't actually tell you exactly how it works. And this is becoming a significant set of issues. More generally, I think there's a worry in a lot of sectors about the fact that increasing numbers of decisions are being made sort of without direct human input. The European Commission in the GDPR actually added a requirement of a right to a human decision over certain things that influence your life. It's hard to know how you implement that, but they're at least worried.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Exactly. That's a tax. It's a cost of doing business. So what I hope to see is actually not a kind of giant action against Google or a giant action against Facebook. What I hope to see is a bunch of kind of small acts of resistance that say, no, this merger, we won't approve because this company might actually sort of grow up to be a competitive threat. And if we do that a bunch, I think we can slowly get our way back to a more competitive market of small startups that get larger, right, that start to threaten the dominance of the whole.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, right. We didn't do anything against the companies. I think the answer is we'll probably see some. We are seeing more activity in Europe. Although the European activity tends to take the form of, exactly. We found you guilty. Pay us $5 billion. Two years later, we found you guilty. Pay us another $5 billion, right? And I think for these companies.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“I think one of the good things about this is I don't think you have to write a new law. Getting anything through Congress right now is a much. But the antitrust laws are written on very broad terms. They give actually a lot of authority to the courts and to the regulatory agencies to make an assessment of competition. And what has been absent is the willingness to be more aggressive in stopping some of these mergers. And, you know, I mean, the problem is, right, you need the political will to do it. Companies both want to merge, of course, right? The VC industry says, yeah, you know, we like this. This is how we get paid. Don't kill the goose that laid the golden egg. But I worry that as we move these mergers increasingly to the incumbents, we are going to end up messing with that.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, you know, I hear that, although I think the idea that the way we're going to fight China is put all of our eggs in sort of one big corporate behemoth and hope that it out innovates China is probably a bad idea. The way we fought Japan, which was the boogeyman of the 1980s who was going to sort of take over all of our jobs and was not sort of let's have one big company that sort of is an American company that competes with Japan, it was a whole bunch of little ideas that turned out to be better and innovating faster.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“So, how do we encourage that? Because one of the arguments that the tech giants make is look, we got to fight China, so we need to be this big. That doesn't matter as much as this fight we're having globally.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, and I think the answer to that has to be we want a bunch of different people with a bunch of different crazy ideas out there in the marketplace because most of them will fail and one of them will succeed beyond our wildest expectations. And suddenly Google will get rocked back on its feet in the same way that Microsoft did when the internet kind of came and surprised it or that IBM did when the personal computer.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Because the sort of PC boom made IBM's dominance irrelevant. But what you can do with mergers is something very different, right? Which is we can stop this problem before it starts. Now, the bad news about that is it doesn't immediately make kind of tech dominance go away. So it's less attractive politically for that reason. But it's something you can do right now and every day that doesn't require an eight-year investigation or any kind of messy breakup or rollback. You could start to say, no, you know what? You're not entitled to buy. Those companies should stay in business on their own, maybe go IPO, maybe stay private. They could sell to somebody else who's not a dominant player and maybe now they'll bring in a new company that has interests in this area.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“I mean, it was successful. I mean, but it was oddly successful. Slowed them down. It slowed them down, exactly, and then other people could come in around them. And IBM was another case where that case went on for 13 years and kind of ended.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, and I think those things can be helpful, but they don't attack the core dominance problem, right? So Google's still going to be the dominant search player. Maybe we're less worried about it if they're not kind of using that leverage in other industries. But I don't think it makes sense to say let's have three Facebooks. Let's just divide your friends group into thirds and say we're going to break them up.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yes, I did notice that. That's just a logo. That's just a logo. And WhatsApp and Messenger are getting more thoroughly integrated to try to make it hard to break them up. But even if we succeeded in breaking those things up and even if Messenger and WhatsApp turned out to be viable independent companies, the market.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Well, so I think one of the advantages of focusing on startups and the acquisitions is it's easier to do this now than it is to try to look backwards and break up something that's been around for 20 years. Sure. So a lot of the political talk is, oh, we should break up the big tech. I think that's hard.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“But a lot of people I talk to in Washington say that they can't keep up as fast as it's moving, that it takes seven or eight years to do these things. And there's not the political will and there's also not the, you know, the FDC people complain. We don't have enough people. We don't have enough ability to do this. Is it gone too far?”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, so we seem to be at a moment in which kind of across the political spectrum in interesting ways, there's widespread agreement that something needs to be done about the dominance of the tech industry. So I think antitrust, which has kind of often taken a back seat to innovation, may be having its moment. And that might be true both in Congress, but also in the agencies and in the courts. So I think there are opportunities to step in and to say, you know what, we're not saying you can never buy another company, but we ought to be looking two, three times and looking pretty closely before we say we're fine with a company buying a startup that are in its space or kind of adjacent to its space because a dominant tech companies have been sort of doing that to preserve their monopoly for years.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“I mean, I think there is some innovation, but part of the question is what's the sector, right? And I think, you know, if the answer is operating systems, they are still dominating operating systems. It's just operating systems are a lot less important than they were 20 years ago.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, I mean, Amazon, Apple, Facebook, Google, right, are the kind of classic companies in the list. Although, you know, it's worth noting that there are companies who aren't in the public conscious who also do a lot of this acquisition in their spheres, Cisco, for instance, has been a real player in this area. Microsoft actually is still doing a fair bit of it.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah. Yeah. I mean, I think it's harder. I think some companies manage it for longer. Google actually sort of held on to a flat market structure inside the business for a lot longer than I thought they would, for instance. And maybe you can do it to the extent that you can. Really, kind of keep it a separate entity, right? And you could probably do that as long as you're not actually a threat to the kind of core business proposition You know, so I don't, I mean, I don't know, I don't have any insight into this, but Whole Foods might do better at Amazon than somebody else who's kind of brought into the company who's a little bit closer to their core mission precisely because they're not close to the core mission.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Right. And so is there any way to stay innovative within these big companies? I don't think there is. I think they get rich. They get ensconced and there's political issues of this exactly happening, things getting shut down or trying small things but not significant things.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“And they're not going to jeopardize that core business for something kind of new and untried. So the problem, I think, with sort of monopolists and innovation is always that most of what you're doing ends up cannibalizing your existing market. Startup doesn't have that problem. They're cannibalizing somebody else's market.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Now, the older eating the young, right? And that, I think, leads us to ossify. We end up sort of staying basically the same because Google has tried all sorts of interesting new projects, right? A lot of people did outside time, a lot of interesting things. But at the end of the day, they have a core way they make 90% of their money.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah. So it's hard to know, right? Because approving negative is hard. It's hard to know sort of what the world would have looked like if we hadn't moved in this direction. I mean, I think one thing you can say is We had more competition even when we had dominant players in the 80s and 90s when we had a broader base of companies. We had quicker turnover.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“With people unexpectedly. Right. And so, in terms of quantifying suffocating innovation, how do we do that? And then the next section we're going to talk about, what we do, what happens to these companies. How do you quantify that? If you don't know what could have been, how do you legally look at that as a society look at that as a government look at that? How do we quantify that? What do you think it's doing to Silicon Valley?”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“The value of that is a couple things. One is just consumer choice, right? It's that, you know what, Facebook pissed me off today. There's someplace else I can go. But it's also the value of kind of diversity of innovation models and ideas. Unless we think we have the best of all possible search engines or the best of all possible social medias, and I don't think we have that, we not only want Google and Facebook to try to improve and continue to make their product better, we want a bunch of other people with different ideas of what makes a good social media network because some of those are going to catch on with people unexpectedly.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, I think that's interesting, right? And, you know, it's, I think probably there are as many sort of reasons for that as there are companies that do that. But they are to me an example of sort of what the world used to have a lot of. Right.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Exactly. And I think that's the other feature that comes in is the people who could go Maverick, right? Who could say, no, I'm going to turn down the acquisition offer, they both have to have that spirit and the desire to do it, but they also have to have the control over the company. And that usually means they are pretty early in the venture capital world, right? If you've gone through maybe three rounds of venture capital, by the time you get to that situation, you may well have diluted your control to such an extent that the VCs collectively have the ability to decide whether or not you're going to sell.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“I think that's right. I mean, every once in a while you see occasional difference. So the great example there is Snapchat. Right, which did turn down an acquisition offer, right? And that didn't end up going well for them”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Absolutely, right? They forestall competition. They get at best, at worst, they get some good new employees, and they have $250 billion in cash sitting around they don't know what to do with, right? I mean, this is literally about the best investment you can make is investing in foreclosing competition or co-opting it, bringing it into your existing monopoly. But it's not good for the world.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Exactly. But I think at least some of them would have thrived outside the environment and didn't make it inside the environment either because we shut them down deliberately or because it's just a different culture and not a thing.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, and a lot of those things get shut down. And a number of people, including Tim Wu at Columbia, have sort of documented all the companies that Google or Facebook have purchased and then shut down. It's really hard to know from the outside whether that was a decision to buy it to shut it down, whether the technology just didn't work and it wasn't going anywhere.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“I think bought because we like the feature, we find it interesting. Well, you know what? The people turn out to be more useful to us over here. Or it's not making money for us in the short run and we are a publicly traded company. So maybe we're going to divert the resources somewhere else.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“It wouldn't sell. Exactly. I think that's right. Yeah, the killer acquisitions are definitely the worst. But I'll say that even the ones that are done in good faith, right, that, oh, I want to add the feature. Well, I mean, so I think it's hard to know because a lot of them disappear, right? But a number of companies that might have sort of had independent existence, Nest labs, for instance, goes into Google Fitbit. Oculus gets bought by VR, Instagram gets bought by Facebook, gets bought by Facebook, Instagram gets bought by Facebook. And some of those companies do quite well. So we've added the feature. But even if they do quite well, they do quite well kind of now as part of the large conglomerate rather than as a separate independent base of consumer choice and power. And some of the companies don't do quite well. So some of the companies, even the ones that”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Right, right. I always say when I read speeches, I said, when's the last time a search company was started? And I was like, Google, it wasn't. When's the last time a social media company was started that would be Snapchat, which Facebook has done its level best to put out of business because it wouldn't sell. Exactly. And when's the last time we had a big commerce player? We haven't. We don't. And this creates an innovation desert, essentially. So what should happen here? I mean, it makes sense that these companies, I don't begrudge them that they would buy these for features. Sometimes like they want to make something better. A lot of times they want to get the talent because there's such a talent war going on, which makes perfect sense. But it's the areas of killer acquisition where they want to kill off a possible competitor or a new innovation that's more problematic, I think.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Exactly right. And I think if it does occur, well, what's going to happen is Google's going to buy you. And as it gets harder to go IPO, as acquisition looks more attractive, even if we had the next better Google, the next better Facebook, they're more likely to get bought up by that dominant company. And so we perpetuate the same company in that position. And if you think about the sort of big tech companies now, they're all 15 years old or older. Most of them are 20 years old. That's an eternity in internet time. I mean, that's, you know, the idea that we've had the same company on top in search for more than 20 years is remarkable.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“So there's nobody that wants to out Google Google, and no one's interested in it. And actually, I just was doing an interview with Ben Horowitz, who he talked about that. And he said, I'm not going to invest in search. I'm not going to invest in social media. I'm not going to invest in commerce. And I said, why? And he said Google, Facebook, Amazon. Why would I? Why would I do that? And that didn't used to be the case. There would be someone that Google wanted to best Yahoo, Blank wanted to best, and that doesn't occur.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“And he said, you know what? Sometimes competition isn't the normal, well, the gas stations are across the street and they have a price war. Sometimes somebody has the market and what you want to do is be the next somebody, right? I'm going to be the next Microsoft. Nintendo's going to be the next Atari. Sega's going to be the next Nintendo. In the same market. Right. Exactly. But sort of replacing it as the leader. So instead of just we have a bunch of companies with equal or low market share, you are the leader, but that prompts a bunch of companies to come in to try to replace you. By making a better product, maybe by coming in kind of in an adjacent or a complementary market that renders yours just not as important.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Yeah, so I think the implications are that we are both driving and then cementing concentration in the tech industry. So there have been a lot of people lamenting the cementing of monopolies among the tech giants. We've had monopolies in tech for a long time, but the thing we've had is what economists call Schumpeterian competition, which is...”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Right. Yeah. Although, I mean, you know, I think with employ right, exactly. And I think with employees, it's always a tough choice because you want those employees to get paid, right? They've often been sort of taking submarket wages and taking stock options. The more successful your IPO is, the tougher a retention issue you have is, right? 100%. Yeah. And so, I mean, I'm not sure. I think if you didn't need the money and you were just making a profit, I'm not sure a company would go public accept that there are a bunch of people who are holding stock options that aren't currently worth anything until they go public.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Let more people invest in private companies, create and open up secondary markets. We've seen some moves in that direction, but I think we could do more to try to say, hey, cashing out as a venture capital doesn't have to mean we're ending the company if we don't hit the billion dollar IPO. It could mean now a new group of investors comes in. Maybe now it's a bank who is willing to loan money, maybe it's a private equity group that does later stage infusions of capital. The early investors get paid and the company continues to operate as a As a going entity.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Or an apple. All right. So when you have an IPO market like that, is there a way to change the idea of keeping these companies private? Is there another way to do this without this idea of focusing on the exit?”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT
“Well, I mean, I think it gets smaller and it gets concentrated with fewer and fewer players. And that's both on the company side, but probably then also on the underwriters and everything else. And I think we drive this continued move towards the easy, the safe strategy being just get bought by Google. Just get bought by Google. And I think to the extent that you now see a bunch of companies starting, right, whose business model seems to be, let's get bought by a Google or a Facebook or an Apple.”
2020-03-09 · Decoder with Nilay Patel · Recode Decode: Mark Lemley · IDENTIFIED FROM THE TRANSCRIPT