POD OF JAKE | #50 - ALBERT WENGER

It's a great pleasure to be here, and it's actually wonderful to hear that you feel it has improved, which would be good. I'd be fooling myself to say I had such a good memory to remember the exact changes, but I loved it the second time through, and I know I liked it a lot the first time too. So let's just say it's improved. I think it was a bit clearer to me the second time around. I don't know if that was just me getting a second read or actually improvements to it, but either way, I highly recommend people go and check it out.

You did a great job of setting it out here at the beginning. Basically, technology comes in these big waves, and when we get a big wave of technology, it changes what the binding constraint for humanity is. The first big technology wave was our invention of agriculture, recognizing that you can put seeds in the ground, irrigate, domesticate animals, and so forth. That led to a change in the constraint from food for foragers—they had to find enough food or migrate or starve—to land: did you have enough arable land to build a successful agrarian society? That was only about 10,000 years ago. Then only a few hundred years ago, with the Enlightenment, scientific breakthroughs, steam, electricity, chemistry, and mining, the constraint changed again. It's no longer how much land you have, because we can make land vastly more productive through fertilizer, for example. But the new constraint is how quickly you can make fertilizer, tractors, roads, buildings, and machines. That's what I mean by capital in the book—physical capital. Today, my thesis is we can make physical capital really fast. A great example is Tesla gigafactories; they just pop these out of the ground at incredible speed, taking old car manufacturers by surprise. Another example: during the coronavirus crisis, the Chinese built a couple of hospitals in a matter of days. So that's not what's really holding us back. It comes down to this attention constraint. I'll leave it there and see where you want to take it, but it really is about these big technological shifts that have been coming more frequently.

That's a great question. The reason that capital is no longer the binding constraint is that all these fundamental needs are not constrained because we don't have the capital. Take Texas right now: people are literally dying because there is no electricity. But that's not because America or even Texas itself doesn't have capital to provide electricity; it's because people weren't paying attention to whether our grid is prepared for an extreme cold weather event and what it would take to be prepared. The thing about attention is: attention is to time as velocity is to speed. Speed is 'I'm going 55 miles per hour,' but velocity is 'I'm driving 55 miles per hour north on this highway going towards the city.' Attention is the directionality of your mind—as time goes by, what is your mind occupied with? You can't go back in time and fix that. Texas can't go back a year ago and say, 'We should have paid attention to what happens if there's an extreme cold snap.' It's like showing up for an exam; if you haven't studied, you can't study at the moment. This is why attention has this fundamental scarcity. The way I define scarcity in the book is through a technological or physical type: something is scarce if it has the effect of preventing us from meeting our needs. People in Texas are dying; their basic needs aren't met, like maintaining body temperature. The reason is not because there isn't enough capital; it's because we haven't deployed that capital correctly, and that's because we haven't paid attention to a fundamental problem. We've entered the digital age with distractions—social media, Netflix—so attention is all over the place. There are problems and opportunities we're not paying attention to, and it's starting to bite us more and more. The climate crisis, the Texas crisis tied to it, is the biggest example of this attention scarcity problem.

That's right. I talk about the job loop, where most of us spend our waking hours at a job to make money, which we spend on goods and services made by other people who have a job. That loop was incredibly successful for a long time, but today we have so much attention trapped in it that could be directed towards other things without a price attached. One example is discovering individual purpose. People think, 'Go to school to get a job,' but there isn't discussion about why you're here. Not many people belong to organized religion, so there's a vacuum. People aren't paying enough attention to their purpose and then have a midlife crisis. So this scarcity of attention exists at the individual and collective level, and it's what we need to work on most, no longer the scarcity of capital.

This is a domain philosophy has had a lot to say about—especially older applied philosophies like stoicism. The question 'What does it mean to live a good life?' is related to purpose. In early industrialization, there was this idea of things yet to be built, and you could see your purpose in helping build them. That was true in the digital realm early on too. There are still many things to be built, but along the way, confusion arose between having a paid job and building something needed. That confusion traces to cultural elements and how education works today. So much of school is about getting into the right nursery school, elementary, high school, college, so you can work at Goldman Sachs or McKinsey. Not enough is built into the educational journey where people are asked to think about a larger purpose beyond themselves, beyond wealth. Related to that, I talk about mindfulness. You can go through school without anyone saying you should develop a mindfulness practice to reflect more freely on things presented through advertising or peer pressure. To disconnect and evaluate on your own requires creating mental space, and mindfulness really helps. We've lost this idea of purpose and substitute it with very industrial age things. It's important to return to making the question 'What does it mean to live a good life?' a common thing for everyone.

I think there's a profound confusion here. We are never going to run out of interesting things to do—spending time with a friend, caring for children, exploring outer space. But whether or not you can get paid for something is a completely separate question. One of the first economists to think seriously about this was Leontief, who remarked how we once had horses used for all sorts of things, and then we replaced them with mechanical equivalents. For many things humans do today, we will be able to use a computer because of the universality of computation. A lot of things we do that we may not think of as computation are in fact computation. A radiologist looking at an image to detect cancer or a fracture is an act of computation. Computers will eventually do them, and in radiology, they're already incredibly strong. So for many things people are currently paid for, it's not clear they can't or shouldn't be. We shouldn't keep a human in a job a machine can do well unless it's what I call 'humans qua human'—something like a live concert or a human server who has tasted the food and can comment. In those roles, we want humans for shared experience. But in all other jobs, we shouldn't strive to keep humans in them just for the sake of a paid job. We need to get more precise: never will we run out of interesting things to do, but we shouldn't force humans to have to do a job to earn a living when machines can do it better, cheaper, faster. The book is very pro-automation, but also pro-creating a system where we all benefit widely from this automation instead of just a few people.

I would start at a different place: most things humans compute today can be automated without needing anything resembling artificial general intelligence. Coming back to the radiology image example, there's absolutely no need for AGI to read those images. Many things people do to earn a living can be replaced by machines that just need to do that task really well. So while I think we can have an interesting philosophical conversation about AGI, I don't think it's relevant to the transition to the knowledge age. The late David Graeber had this great expression 'bullshit jobs'—many jobs where people are basically unhappy and stuck, and society doesn't give them an out. These could or should be automated. Every time we hit an economic crisis, the labor market doesn't recover as much because companies roll out technology to make up for it. For example, after the 2008 financial crisis, investment banks and insurance companies downsized and aggressively deployed technology. Now, on creativity, there are creative jobs, but base-level creativity—like coming up with a new dress pattern or color scheme—we already know generative adversarial networks can do a fantastic job at those.

Tasks. So I do think there are higher levels of creativity where, now we're getting more into the question of artificial general intelligence. But let me give you even another example where I just don't really think we need artificial general intelligence, which is, for instance, customer support chatbots. They've made a huge amount of progress because generally it's a very domain-specific thing. You're dealing with a specific system, let's say your internet service provider or your insurance company. The machine doesn't need to be able to have a conversation with you about philosophy, love, or God or whatever else. It just needs to be able to locate the right insurance form or answer the question of what goes into a field. So I think I love the speculative part of talking about artificial general intelligence, but it's kind of somewhat disconnected from our ability to automate vast swaths of current human activity.

The knowledge loop is why we're surrounded with all this knowledge we've accumulated. It's simply this ability where you learn something, then you use what you've learned to create something new, and you share that new thing. Now somebody else can learn from that thing. That loop has always existed, but we can use digital technology to turbocharge that loop. YouTube is a phenomenal case study, both the internet at its best and at its worst all in one place. At its best, you can watch videos from Veritasium or Numberphile, or learn how to play guitar or garden. You can easily share your own insights back to others. But it's also optimizing for maximum attention because their model is to sell ads. How do you get maximum attention? You create algorithms that optimize for engagement, and engagement is often driven by emotion, which is often driven by outrage or upset. Without saying that Google started out to be evil—they certainly didn't—but if your model brings you in a direction where you're trying to hog attention, your algorithms will optimize for misinformation instead of giving people access to this type of knowledge. Interestingly, because I've watched enough Numberphile and other videos, the machine now happily presents more of those to me. But I think it's easy to get pulled into very different corners of YouTube by the same algorithm.

One of the important ideas of the book is that when you have a big transition happening, I'm not suggesting that I know what the knowledge society should look like. Instead of designing the knowledge society for you, the book says how can we give people more freedom so they can figure out how to organize. I talk about three freedoms: economic freedom, informational freedom, and psychological freedom. Economic freedom is some form of universal basic income, decoupling your ability to live from the need to have a job. In my mind, these are relatively small amounts, maybe as little as a thousand dollars a month per adult and less for children. This isn't about living in Manhattan; you might have to share a very small apartment if you wanted to make that work. But it's enough so you could decide to live in a rural area and explore things. The reason I love this model is because it lets us embrace automation, and it doesn't prevent anybody from working—you can always earn more money if you want. But you can have many different ways of doing that. Informational freedom has to do with who controls computation. I think this is an area where we're making a huge number of mistakes from a policy perspective. So much computation is controlled by just a handful of companies: Google, Amazon, Apple, Facebook. The issue is that we all have supercomputers in our pockets. A smartphone is a supercomputer that can talk to all other supercomputers in the world and all the server computers. Yet your ability to program this phone is extremely limited. Yes, if you're a programmer you can write an app and get it distributed through the app store—important caveat: app stores can take apps down any moment, so there's a chokepoint right there. But you can't actually innovate on top of the apps already on your phone. You can't take the Facebook app or the Twitter app or the Amazon app and script them. They are reducing you on your supercomputer to an idiot with your thumbs trying to do stuff. If those things were scriptable and programmable, somebody else could innovate on top of them. A lot of the informational freedom section is about how we can make sure computation and information aren't controlled by a few large entities and maybe the governments. Psychological freedom is about the idea of mindfulness. Our brains didn't evolve in a world where if you saw a cat or a fox, there was an actual cat or fox. Now I can scour the internet and provide you an infinity of cat pictures. It's very easy for these companies trying to hog our attention to trigger these emotions—cute, upset—and use them to hog our attention. Psychological freedom is about how, if you had economic and informational freedom, you can truly be free from the onslaught of all this stuff around us. I compare it to sugar. Our bodies didn't evolve in a world full of sugar; we like sweet taste. Now we can add sugar to everything, and the commercial interest is to add sugar to everything so you eat more. We have the same relationship to information and computation at the moment. The good news is we can use mindfulness to overcome it. These freedoms interact with each other. If you have economic freedom, it makes it easier to use your informational freedom because you can decide to write open source software without needing to make money. Psychological freedom makes it so you might not be embarrassed sharing something half-baked, and if criticism comes along you won't recoil and stop sharing. These three freedoms build upon each other, creating the ability to invent the new. People will take very different cracks at what the new looks like, but we have to give them that room instead of keeping them trapped in the old. Too much of politics is about trying as hard as possible to patch the industrial age, keeping people trapped in it, instead of letting them figure out what this new thing might look like.

Yeah, I think this is a super important point. Needs are things you have to have to survive. If I were in a spaceship by myself, what are the things I need to survive? They aren't just bodily functions; they include purpose. You could be in a spaceship surrounded by enough food, but if you were there by yourself and didn't know what mission you were on or for what reason, you'd go insane and stop eating. When I say survival, I don't mean just physiological survival; I mean a complete survival of being able to live my life. This is important because when people look at these small amounts and this notion of meeting needs, they say who wants to live like that? It's subsistence. What they don't recognize is the extraordinary freedom it provides once those are taken care of—the freedom to explore almost anything else: artistic and diverse knowledge endeavors, other forms of work. You can only have the freedom to explore those once that foundation is set. We have confused our needs with our wants. There are lots of things people want, but they aren't necessary. I'm not giving anybody grief for wanting something; it's just being clear about that distinction in your own mind, because then you can decide if it's worth working harder for it or if you don't actually want it. I give some examples to make this less abstract. People say, 'What about skiing, Albert? You love skiing.' Skiing is incredibly expensive, so how can anybody on basic income afford a lift ticket? I say that's not the only way to ski. People skied long before lifts existed—they hiked up the mountain and skied down. That turns out not to be expensive. I know people who do that, living in a mountain village where they can do a lot of skinning. They're not heli-skiing in some monstrously expensive way. There's a lot of confusion about how to do things, fueled by tons of advertising saying you need the big car, the big house. Trillions of dollars over the last 50 or 80 years have been spent making everybody confused about what they need versus what they want. Once you re-establish that as a mental discipline, it provides extraordinary freedom.

Absolutely. I think it's crucial for us to have positive views of what can be achieved. I'm very optimistic about what can be achieved with technology. When I say technology grows the space of the possible, the first example I give in the book is fire. Fire is the first human technology. It's great—you can use it to cook, but you can also use it to burn down somebody else's village. Anything we've ever invented has included positive and negative use cases. That's what I call the space of the possible—it gets enlarged by technology. I gave YouTube as a great example: sharing and watching YouTube has many positive use cases, but scaled manipulation is also included in that space of the possible. I think it's incredibly important to come up with narratives of the good things we can do. Often they unfold as the technology unfolds. Some people today say you should already know in advance that this thing is going to be bad. Clearly, there are some things you can know might not be a good use of technology in advance. But as far as technology itself goes, take facial recognition, which is heavily debated. My view is it provides capabilities that include very positive and very negative use cases. A positive use case: if a person is lost, being able to find them would be incredibly positive, and that might require street cameras. We need to think about how to construct a society where positive use cases far outweigh negative ones. Narrative really helps. I believe the lack of a positive narrative, a future that people find exciting to work towards, has given undue power to people like Trump and backward-looking narratives that say the answer is to go back to the past. Positive utopian narratives—I like to spell utopia with an 'eu,' which means 'good place'—are incredibly important as motivation, pointing to a future worth striving for.

Sure. You can find the book at worldaftercapital.org. I am still working on it, making some changes, but you can download a PDF and print it for yourself easily. You can also follow me on Twitter, where I'm just @albertwenger.