Moore's law (E)

0:00Gravity isn't simply a good idea. It's the law. Hey, this is Emily in the Bronx, and you're listening to a special archived episode of Akimbo.

0:18Before we talk about the most important law of our lifetime, it's worth taking a second to distinguish between a law and a theory. In popular culture, a law is bigger, better, more powerful than a theory, but in fact, the opposite is true. All a law is, is a statement of how things happen, of how the world is. A theory is our explanation as to why it's happening. And theories get better over time because people come up with better, more accurate explanations of what we're seeing. Theories can be tested. Laws are simple statements of how the world is. Douglas Engelbart was one of the most important pioneers of the computer world as we know it today.

1:12In 1968, he gave a demo. And for people who were paying attention, it was the very first time they saw the mouse and networks and hypertext all at once, one guy inventing our future. But about 10 years before that, Douglas wrote a paper. And in that paper, he speculated about the fact that computers were going to get smaller, faster, and cheaper. This is in 1959, a lifetime ago. Well, about five or six years later, Gordon Moore, who worked in the semiconductor industry, was asked to write a paper, a short paper, predicting the future of computers. And in it, he posited a theory. And his theory was that based on innovation and progress that was getting made in certain kinds of production, computers would get faster and cheaper.

2:15In fact, every two years, the same amount of money would buy you twice as much in terms of computing power. One of his colleagues at Intel, where he ended up working, David House, turned it into Moore's Law. And what he said was that 18 months, every 18 months, computers double in power and speed and efficiency. That simple law, which is generally true and has been true for my entire life, is responsible for the world we live in now. If Moore's Law applied to your house, you would be able to buy a house as nice as the one you live in for$30. Because unlike almost everything else in our world, computers keep getting faster and better and cheaper.

3:09It's worth understanding why. Because just as you can, with effort, fight gravity, you can, with effort, build a life that goes against the magic of Moore's Law. But understand that you're going to have to work at it and work at it, and it's going to get harder and harder. When I was growing up, I read a book by David Brin called The Practice Effect. And my memory of it goes like this. Imagine a world in which things got better the more you used them. That if you picked up a stick and started using it as a crutch, that every day over time, instead of it getting weaker or more scuffed, It would get more and more like the best possible crutch.

3:57That entropy worked in reverse. That practice made things better. And so in the novel, he theorized that people who didn't have a lot of resources would work for rich people using their stuff. So that when you got home, your knife would be extra sharp and your frying pan would be extra nonstick. And everything about the world you lived in would keep getting better. because if you left it aside or put it on a shelf and didn't use it, then it would start to decay. Well, that's a law that's not true in our lives, but Moore's Law has been true. So here are some theories about how it will continue because it keeps broadening.

4:39It's not just that the CPU in your computer gets faster and cheaper. In fact, so many things that computers are touching are affected by the factors that keep making it better. So let me highlight three of them. The first one is the idea of the experience curve. Years ago, LED watches with their little lights on the front showing you in numbers what time it was, cost between$1 ,000 and$5 ,000. dollars. They were clearly a novelty. Texas Instruments decided, after doing some analysis, that they would break open the market. And so what they did was they priced their watch at a price below what it cost them to make.

5:28They lost money on every one. Because it was so much cheaper than the other LED watches on the market, their volume went up. And as their volume went up, their production capability got better. And as their production capability got better, it got cheaper and cheaper to make a watch. And at that point, even though their watches were$1 ,000 less than the competition, they were now making a profit. Because the experience curve says that the more we make something, the better we get at making it. And this is particularly true with anything that involves data. That once you build a database or a silicon chip or the stencil for a memory chip, the next one you make is cheaper still.

6:14That once you build a fab and figure out how to make it efficient, each one gets cheaper as it comes off the line. It's worth noting that this is not true for things like drilling oil out of the ground. Because as you get rid of the easy oil, the hard oil, deeper down, is harder and more expensive to get. But as we start working with ideas, and as we start working with processes, we discover that the opposite is true, that the experience curve makes us more efficient at scale. Number two is that computers can now make computers. This is a huge leap forward, and I don't think Moore talked about it in his original paper, but what it means is that as we get more powerful computers, those more powerful computers are helping us make more powerful computers, accelerating the curve forward.

7:11And the last one, and the last one, the last one is super important and brings us back full circle to Douglas Engelbart's 1968 demo, which is the network effect. What we know is that big networks get more efficient. What we know is that computers by themselves are sort of powerful, but computers networked to one another are incredibly powerful. And so if we apply the math of House's analysis of Moore's Law, we see that, yes, it does keep doubling every 18 months because the network effect pushes it forward. So what does this have to do with you? You're not busy working with Carver Mead and designing fancy computer chips.

7:58Of course, it goes beyond your fictional house getting fictionally cheaper. The real examples are pretty stunning, and you don't need to be a genius to figure out what happens after that. The computer you could have bought in 1979, with the money it cost to buy the smartphone in your pocket, didn't even exist. A couple of years later, you could have bought a computer probably the size of a packing crate that had one hundredth the power of the computer in your pocket right now. So what happens 20 years from now when the computer in your pocket doesn't cost$800 but costs$8? What happens when the network effect continues to create more and more value for the people who create the networks?

8:54What it means is that the tires in your car are going to have computers in them, that your clothes are going to have computers in them, that everything around you that you touch, that you see, that you interact with, at some level is going to be network-enabled. So what happened at Yahoo? When I was at Yahoo, Google was a tiny, tiny little figment of someone's imagination. How is it then that in less than 10 years, they won and Yahoo lost? Well, what most people don't remember is that Yahoo wasn't a search engine. It was a directory. There were hundreds and hundreds, in this case it's worth using the word literally, there were literally hundreds of librarians sitting in a room in Silicon Valley surfing the web, one website at a time, figuring out what belonged where in the directory.

9:53because, of course, a computer couldn't do that. And the guys at Google redefined the problem and turned it into one that was brain-dead stupid, that was so stupid and simple that a human would go crazy trying to solve the problem and you couldn't possibly hire enough humans to use PageRank to organize the web. But you could hire enough chips to do it because every day they get cheaper. However, the idea that we are living in an economy shaped by something that's about the opposite of economy, scarcity, which is abundance, more computing power tomorrow than we have today, more network effect tomorrow than we have today, means that just about everything we do is going to be impacted by Moore's law.

10:48because the theory shows us that even after the chips have gotten pretty much as far as they can go using existing technology, we're going to have to either show up with quantum computers or something else to keep the chips going faster. The things outside the chips, the fact that we live with the experience curve making things better and the network effect making things better means that Moore's Law isn't going to get repealed anytime soon. And so the question is, are you acting like Yahoo and insisting that a pre-surplus computer mindset will persist? Or are you betting that the stuff that you use, whether it's video, online connection, computing power, the Internet of Things, will keep getting better and cheaper, thus making your work better and more efficient.

11:47So we've got to pick which side of the fence do we want to be on.

12:13Berta Perry. My question is... And that completes my question. If you've got a question about this week's episode, we would love to hear from you. Visit akimbo.link, that's A-K-I-M-B-O dot L-I-N-K, and click the appropriate button. Hi, Seth. Andy here. Thanks so much for all you do. question about exclusivity and merit. So if we have a Facebook chat where lots of people end up joining in, it becomes really diluted or even in the physical world with a tea party or birthday. So how can we maintain intimacy if we need to limit and exclude folks? Then we have to pick and choose rather arbitrarily.

13:01And I hate that. So is there another option? Or if we have to be exclusive, how would you recommend designing an organization or gathering? So it's true, of course, that thanks to the network effect, groups keep getting bigger, but big groups get less efficient. It's almost impossible to have interactive one-on-one conversations with hundreds of people at a time. As a result, due to the sheer physics of information flow, we have no choice but to create smaller groups, which means we have to choose who is going to be in the room. The important thing to do is to acknowledge that we're either choosing randomly or semi-randomly or on a particular form of merit.

13:54So if, for example, we're allocating people to lifeboats, the only thing that matters is how much you weigh. Because once we have the right weight in the lifeboat, off you go. It's been said that at the most famous colleges, they eliminate 70 or 80 % of the applications based on their take on merit. And then after that, it's random. And that's probably true. So if you get rejected from college, you need to be able to say to yourself, well, I might have been qualified to go, but at some point, because there's only 140 slots, the last tranche, the last thing I have to get through is random. And I think we can have the confidence to say to other people, this group only has room for 110.

14:44They're not the 110 tallest people or the 110 people who play the ukulele the best. They're just our 110 people, and that's okay. Because scarcity creates value. Access can't be open to everyone, everywhere, all the time. But at least we should be honest about how we are making our selection and not pretending that there's some sort of magic metric, not keeping it a secret, and not using a metric that doesn't actually line up with what we're seeking to do. Hi Seth, it's Eric from Orlando, Florida. After listening to the last episode, I realized that trying to maintain a good GPA is trying to seek marriage from a system based, at least partially, on compliance.

15:33And this is because that part of my grade in some classes is for attendance. Even if I have a bad professor, I still go to lecture for the grade. So my question is, how do I get rid of my habit to comply even though I know I'm in a system that rewards it? Thank you for your thoughts. I feel your pain and thank you for this question. Mandatory education can be a bear because it's not voluntary, because you are a part of a system that you didn't choose. So let's start with that analysis first. You're in a system you didn't choose, but if winning in that system sets you up to get to the next level of where you seek to go, then that might be a price you are willing to pay.

16:20If losing in that system doesn't matter to you, then go ahead and lose, but be prepared for the consequences. Lots of us deal with systems we didn't choose, and so there are consequences to our actions. You might not like working out in the gym, which is fine, except then you don't get to walk around like Linda Hamilton or Chuck Norris. You're going to have to deal with the consequences of that, even though you didn't choose your metabolism. So the same thing is true in school. But once you get past 12th grade, you're in a system you chose. So if you're in a system you chose, the question is, why did you choose it?

17:02And if you chose it because it's a ticket to get to where you want to go, then it helps to remind ourselves that we chose it. So we can say, I think it's stupid that I'm going to get judged on my grades in physics because I have no desire to practice physics, but I am going to be judged on my grades in this physics class. And since I chose this path, that's a price I already decided to pay. You don't need to reconsider the choices because you already made them. Hi, Seth. This is Kyle from Ecuador. In your episode, Interoperability, I think the story can go one level deeper. The listener might conclude that regulations to break up Standard Oil or prevent tie-on of beer companies is always good.

17:53However, the reason Standard Oil was able to get rebates on rail prices in the first place was because, you guessed it, government regulations. The government required all rail traffic to be treated the same and charged the same price. Thus, government created laws that enabled standard oil to create a monopoly. But that's not all. Going one level deeper, it's not clear if the monopoly was bad. Yes, many oil businesses were unfortunately strong-armed into selling to standard oil. But the consumer was very happy because oil prices fell from 59 cents to 7 cents a gallon due to improved efficiency.

18:34This would be like if your monthly phone bill dropped from$70 a month to$10. There's also more to the story on bars. As a former bar owner, I can say that government regulation is the reason why it's so expensive to start a bar in the first place. It can cost up to a quarter million dollars just to get a liquor license. a regulation that provides no benefit to the consumer. If we got rid of liquor licenses, there'd be no need for beer companies to offer tie-ons in the first place. Regulating tie-ons is just putting Band-Aids on top of Band-Aids. Maybe we should remove the Band-Aids suffocating the problem and let the wound breathe and heal.

19:14Allowing utilities to charge different amounts to different kinds of traffic makes a lot of sense, actually. Right now, there's a big push to enable electricity utilities to charge more at peak times and less at non-peak times. This would make solar power more viable, as power would be cheap when the sun is shining and expensive when it's not. Similarly, if one person is backing up their videos while another is trying to provide time-sensitive advice on surgery via VoIP, it obviously makes more sense to prioritize the traffic of a time-sensitive surgery and push the lower-priority video backup to the middle of the night when there's less demand on the network.

19:54And thank you for this question about interoperability, an episode we did a little bit ago. There are good arguments to be made for industrialism. It has been pointed out, for example, that Walmart is an economy that's bigger than almost every country in the world. and there's almost no graft, there's very little violence, and it's extremely efficient. Hard to say that about many countries in the world. So the argument goes, bigger can be better because industrialists have the power to put in place efficient systems. And there are two challenges to this. The first one I think is the most important, which is choice.

20:38Choice matters. Choice is important. Choice makes us believe we have agency over ourselves and our future. Choice is at the heart of what it means to be in a free culture. You can do this or you can do that. As we just heard, there are consequences to your choices, but there is choice. And what happens once a company passes the threshold, and Walmart is not at that threshold, once it passes the threshold to become a monopoly, then the incentives of the monopolist are different. They come out ahead when two things happen. One, when they keep their monopoly power and they do that by keeping innovation away, by keeping alternatives away, by using their leverage and their cash to change the culture and the government so that no one challenges them.

21:35and when they take choice away from the consumer. Because if the consumer has no choice, the profit-seeking, profit-maximizing, short-term-focused monopolist will take advantage of the fact that the consumer has no choice. And so when you have this no choice, Hobson's choice, as it's called, you're not happy. You're not getting the best that's available to you. You don't have freedom. Taking away choice is one of the problems of monopoly. The second challenge, which is related to the first one, is that when innovators, when entrepreneurs, when people who want to make things better, see a monopoly operating at full power, they tend to decide on their own to find another place to be.

22:30That if you were looking to make a mark in the 1960s, becoming a musician might have been tempting because it was far from a monopoly. Lots and lots of people could pick you. Someone like Barry Gordy could start a record label. It wasn't easy, but he could start a record label that would change the culture. It wasn't completely locked down. That the difference between the world when AOL was around, where you had to have a meeting with Ted Leonsis and get a green light in order to start a content site on the thing that felt like the internet. And what happened just a couple years later, when the internet said, you want to start a site, go ahead and start a site.

23:15The difference in innovation was stunning. And so as a culture, we benefit from when there's permeability and interoperability. We benefit when many people believe they have a choice, not just a choice as a customer, but a choice as a creator. So I'm not arguing in the interoperability episode that every company needs to be tiny. I am well aware that the culture we live in is largely possible because 50 giant corporations have woven together a stable, efficient net that lets us get what we want when we want it. and they do it in a way that makes people feel free, like they have a choice, like they have leverage, and like they have power.

24:04My argument is that if Standard Oil, I kept doing what Standard Oil was doing, we would have been locked in a downward spiral when it came to things like energy. And once they had cornered the market on energy, they wouldn't have stopped. Because monopolists understand that the Achilles heel of capitalism, the thing that makes it all fall apart, is when we take away choice from the consumer. Thanks again for listening. We'll see you next time.