⚔ The FTC's Lina Khan vs. GenAI
⚔ The FTC's Lina Khan vs. GenAI
Also: 5 Quick Questions for … aviation expert Jon Ostrower on the challenges of supersonic commercial flight
Quote of the Issue
“They will race behind you, they will stumble, they will fall. But in time, they will join you in the sun, Kal. In time, you will help them accomplish wonders.” - Jor-El, Man of Steel
The Essay
⚔ The FTC's Lina Khan vs. GenAI
You have to dig a little to find the thesis of FTC Chair Lina Khan’s essay in The New York Times, “We Must Regulate A.I. Here’s How.” It’s way down in the twelfth paragraph, almost near the end: “The history of the growth of technology companies two decades ago serves as a cautionary tale for how we should think about the expansion of generative A.I.”
There’s your trouble. The story of Big Tech in the 2000s isn’t a cautionary tale, not at all. Why would it be? These companies are the crown jewels of the American economy. Europe would love to have them, and China is trying to build them. To answer that question, Khan offers a version of the “surveillance capitalism” critique of the digital economy created by companies such as Facebook and Google. From her essay:
Those innovative services, however, came at a steep cost. What we initially conceived of as free services were monetized through extensive surveillance of the people and businesses that used them. The result has been an online economy where access to increasingly essential services is conditioned on the widespread hoarding and sale of our personal data. These business models drove companies to develop endlessly invasive ways to track us, and the Federal Trade Commission would later find reason to believe that several of these companies had broken the law. Coupled with aggressive strategies to acquire or lock out companies that threatened their position, these tactics solidified the dominance of a handful of companies. What began as a revolutionary set of technologies ended up concentrating enormous private power over key services and locking in business models that come at extraordinary cost to our privacy and security.
Sounds pretty serious. And it must frustrate those who worry about “surveillance capitalism” that there’s little evidence most of us share their concern. All those “invasive” business models that provide services at “steep cost” — in the words of Khan — seem to provide an acceptable tradeoff to most of us. This view was nicely summed up back in a 2019 piece by Axios reporter Erica Pandey who wrote, “I, like scores of others, have decided that I’m OK with giving up personal data in order to keep getting convenient, cheap (or free) services. Despite the known episodes of firms misusing data, the ease and quality of life under the reign of Big Tech generally seems worth it.”
Indeed, many of us would need substantial compensation to give up our favorite online services. In the working paper, “Using Massive Online Choice Experiments to Measure Changes in Well-being,” economists Erik Brynjolfsson, Felix Eggers, and Avinash Gannamaneni use a survey method called willingness-to-accept, which asks how much money people would demand to stop using a good or service for a certain period of time. The study finds that Facebook users had a median WTA of about $40-$50 per month, meaning that they would need this amount of compensation to leave Facebook for one month. The study also finds that search engines, email, and maps were the most valued categories of digital goods, with median WTA estimates of $17,530, $8,414, and $3,648 per year, respectively. Video streaming services, such as YouTube and Netflix, had a median WTA of $1,173 per year, which was much higher than what consumers actually pay for them ($10-$20 per month).
“Our measure suggests that the surplus the median consumers receive from these goods is a 5-10 multiple of what they actually pay,” the economists write. Other categories, such as e-commerce, social media (excluding Facebook), music, and instant messaging, had lower median WTA estimates, ranging from $842 to $155 per year.
A total misunderstanding of revealed consumer preferences is a big part of the problem with “surveillance capitalism.” An even bigger problem is a fundamental misunderstanding of the online ad business. A great explanation from Stratechery’s Ben Thompson via my 2020 podcast chat with him. Thompson said he didn’t much like the term “surveillance capitalism” as it suggests a vision of companies accumulating personal data in a manner similar to the East German Stasi. Thompson:
The way this data is actually accumulated and used is a series of vector equations that are inscrutable to anyone. And when you go in, that data is not sold nor processed. What an advertiser does is they walk to Facebook and says, “We have a product that really appeals to men in their late twenties. Primarily we want to target men in their 20s in Los Angeles that likes sports.” And Facebook doesn’t say who those people are. They just say, “Okay, we will find people that fit that description, and we will put this advertisement in front of them.”
Thompson went on to say that this system is actually positive for economic growth and small business formation. By allowing small businesses to advertise directly to their intended customers at a lower cost, the internet has enabled a wider array of services to be offered to consumers. This is in contrast to traditional TV advertising, which is often more expensive and less targeted. The benefits of accumulating data are often overlooked, and privacy laws can end up cementing the position of the biggest players in the market.
Unsurprisingly, Khan’s devotion to the notion of “surveillance capitalism” gives me little confidence that a Khan-helmed FTC is ready to wisely regulate AI.
5QQ
💡 5 Quick Questions for … aviation expert Jon Ostrower on the challenges of supersonic commercial flight
It’s been a half century since the FAA banned supersonic commercial flights over the continental US over concerns about sonic booms. In the 50 years since, “airplane manufacturers have innovated on margins other than speed, and as a result, commercial flight is safer and cheaper than it was 40 years ago,” note the authors of a 2016 Mercatus Center paper. Those are laudable improvements, but I suspect the future-optimists of the 1960s would be shocked to hear that passenger travel today is no faster than it was in the Kennedy era. Fortunately for all you speed junkies, startups like Boom Supersonic are working to develop quieter supersonic planes that may one day bring us an age of affordable, supersonic air travel over the US.
Still, FAA requirements are far from the only challenges facing the engineers at Boom. As The Air Current reported late last year, Boom has faced some difficulty securing engines for its new SST design. And of course, there’s always the economic problem: Faster planes burn more fuel, which drives up costs. To get a better sense of what’s going on, I tapped aviation enthusiast Jon Ostrower.
Jon is Editor-in-chief of The Air Current, which reports on all things air travel, including topics that may interest Faster, Please! readers, like electric aviation, air taxis, and more!
1/ Is the biggest bottleneck right now for commercially viable supersonic transport a technological bottleneck or it will prove to be a regulatory bottleneck?
Yes. It really is a combination of both of them. I think you need a technological breakthrough to begin to have the regulatory conversation. If you can show from a technological perspective that you can, number one, mitigate the supersonic boom that comes along with crossing through the speed of sound, that is going to change the conversation. Because all of a sudden you've solved that problem. The second piece of that is the cost of operation. It's an economic question as well: Can you operate these aircraft sustainably from a profitability perspective? And if you can't do that, in conjunction with solving the regulatory question, which is the supersonic sound profile, they have to really go hand-in-hand together.
2/ There are several startups out there. How close are they to solving that problem?
Right now, the closest we've gotten to getting rid of the sonic boom, which would ultimately enable over-land flying — the benefit there is not just the ability to fly over land, but you can just open up the number of routes that you can fly viably. If you're barred from a market, it inherently limits the business case. So in the case of actually reducing the sonic boom, the closest we've gotten right now is a demonstrator program coming out of NASA and Lockheed, and it's called the X-59. It is designed to reduce the sonic boom to what they've described as effectively the closing of a refrigerator door rather than a rattling of your entire home. They think that's a major obstacle that they need to get through to the other side of it. And so with the public funding for NASA that comes along with this, ultimately the results of the testing, depending on the structure of the program, would be available for industry to use and learn from and take it from there. This is just a demonstrator program. It's a single-seat test aircraft. It sort of looks like a lawn dart. But it is a big step in that direction to try and understand what causes, and what can ultimately be done to reduce significantly, a sonic boom.
3/ What are these startups like Boom doing? What’s their plan for how they will make supersonic flight work?
Boom, from a commercial perspective, is the furthest along in terms of its establishment of a sort of commercial enterprise to actually go do this. Their assumption is that they're not going to be able to fly over land. Their routes are going to be limited to over water. And they've said there are about 600 routes in the world that can be supported by a supersonic transport at the size and the economics that they're looking at. Notably, we have not seen the list of 600. There is a very significant debate there about the number of routes that you can use this on. And I think one of the really interesting historical pieces here is that when Concord was in operation, Air France and British Airways never flew against each other on the same route—there was never a supersonic plane versus a supersonic plane in the market—so we never got to see what that battle might look like in terms of how supersonic economics affect one another in competition. And I think that that's a really important piece of understanding the larger market for any more than 12 supersonic airplanes that were flying with Air France and British Airways over the years.
There are a lot of economic benefits that come with reducing the amount of time between points, like you can have one less crew member in the flight deck and the flight deck wouldn't have to take a rest. So that would be a demonstrable savings. Yes, you'd burn way more fuel, but you'd be able to, in theory, offset that through other revenue premiums in terms of speed and all that. What they found, as they did a kind of war game scenario on what this looks like in the market, was that a supersonic airplane flying against a supersonic airplane ultimately competes on price and competes for the same passengers and the same tickets. And so, the subsonic airplanes would've to cut their price even more to offset the premium that you'd get from speed. What Boeing ultimately concluded, one of the reasons that their customers were not too particularly happy with the idea of them of pursuing it (aside from the fact that it would've burned more fuel than sort of the expected step change that you'd expect from a new airplane), was that you'd actually take revenue out of a city pair because of the competitive dynamics against subsonic airplanes.
So it's a fascinating economic equation as far as, can airlines deliver a revenue premium on supersonic? And again, we've never seen them competing against each other, so we don't really have that data. But in terms of the overall thinking on this, it does actually disincentivize speed.
4/ If you go back to the early ‘70s and the US makes a different decision, letting airlines fly supersonic across the country, what does the industry look like today?
It's a great question. I guess the next question is, does deregulation happen at the end of the ‘70s? That's the big question, because going from New York to L.A. in significantly less time—there is something really attractive about that. That's the like sexiness of this whole thing, right? Let's play this out: In ‘71, they canceled the program; in ‘73, they banned overland flying with supersonic. Had things progressed, it probably would have been prohibitively expensive for the airlines to operate in a deregulated environment. When the Civil Aeronautics Board was willing to operate with a guaranteed return on invested capital for the airlines in terms of setting ticket prices, that was great. You could totally run that all day long.
But with the overall sort of liberalization and opening of the market to carriers like Southwest, which actually predated deregulation—this sort of free-for-all that came into play in the ‘80s—the big question has always been, what do airlines want? Not just, what do passengers want? What do airlines want? And airlines ultimately want to be able to deliver their passengers from point A to point B for the lowest possible unit cost. And the question is, is there a revenue premium that you can get for speed? Certainly when it comes to business aviation, yes. Absolutely. You've got Gulfstreams flying at north of 90, 91, 92, almost 93 percent of the speed of sound. That's really fast. And obviously, the prestige that comes along with that comes with a $75-$80 million price tag. There are customers that will pay for it.
In terms of the airline service, would we have come down the cost curve? Yeah, we absolutely would have, because the natural evolution of technology and the necessity for those lower costs would have ultimately caused engine makers and aircraft manufacturers to refine that to be more efficient. The question is, when you roll to the early ‘80s and the 767 comes along with significantly lower unit costs, it's got one less engine than the DC-10, two fewer engines than the than the DC-8, and it's flying cross country and it's burning just way less fuel and airlines go, “Okay, is there a benefit for speed?” Maybe not. Fast forward to 2023, and we're 10-15 years into in-flight internet and all the other things that make speed less important. It is a fun counterfactual, though.
5/ I was looking at the 1986 Reagan State of the Union speech where he talked about creating a new Orient Express, a 25x-the-speed-of-sound airplane. Have you ever written about that?
It’s interesting. Reagan was a huge aerospace proponent. Huge. Obviously with the Freedom Space Station and Star Wars from a space perspective. But as governor of California, Lockheed and McDonnell Douglas were in his backyard and he was, as governor, at the rollout of both of those airplanes. He was keenly aware of what commercial aerospace could do for an economy and ultimately where we kind of hoped it was headed. And it is aspirational. It’s exciting to think that at a time when we were trying to bleed the Soviet Union dry on Star Wars and other defense spending, you really had this push where the Soviets were trying to do supersonic — also even before that in the ‘70s when we had Concord and the Boeing-Lockheed programs with the government. Twenty-five times the speed of sound is pretty darn impressive. The Space Shuttle only goes 17x. How much faster do you need to go?
Micro Reads
▶ Google DeepMind CEO Says Some Form of AGI Possible in a Few Years - Miles Kruppa, WSJ |
▶ Fusion and the Holy Grail - Tristan Abbey, The New Atlantis |
▶ What does it cost to build nuclear power? - Brian Potter, Institute for Progress |
▶ Reactor experiment demonstrates alternative fusion scheme - Daniel Cleary, Science |
▶ Biden’s vaccine project needs to be more like Operation Warp Speed - Alec Stapp and Arielle D'Souza, WaPo Opinion |
▶ Microsoft Economist Warns Bad Actors Will Use AI to Cause Damage - Bryce Baschuk, Bloomberg |
▶ Elon Musk Tries to Direct AI—Again - By Berber Jin and Deepa Seetharaman, The Wall Street Journal |
▶ This Robot Can Paint. But Is It Art? - Oliver Whang, NYT |