For most products, the digital UI is like the DMV. It’s designed for everyone, so it works for no one.

A huge percentage of the tasks we perform in software are lots of clicking, moving the mouse, waiting, and typing, simply to perform something which could be described comprehensively in a single english sentence. Let’s call that 60% of software action. Within the remaining 40%, there’s another 60% which has a few more steps and inputs to it, but 

these steps and inputs do not necessarily map well to existing user paths in the target application’s UI.

 That is to say, every user has their own context and situation which demands a different UI to satisfy their needs.

Consider what a user does today when he is ready to Uber back to his hotel. It goes something like this:

Open Uber app. Tap search bar. Try to remember the name of my hotel. Close Uber app. Open email app. Scroll through emails to find reservation info. Can’t find it. Email search bar. Type ‘hotel.’ Scroll past 2.4 million hotel ads. Find reservation. Mentally note the name of the hotel. Open Uber app. Tap Search bar. Type name of hotel. Tap the right one. Continue tapping buttons until Uber finally acknowledges that someone is coming to take you to your hotel.

 “I’m ready to go back to my hotel. Uber Black please.”


 “Ok. Here’s a picture of the car that Uber found for you. Your driver has a 4.8 rating and the content of the reviews are solid. His name is Hasan, and he will be arriving in 4 minutes.”

Nearly all UX sucks. Here’s another one. Let’s say you just remembered that your father in law is coming to town, and you want to make sure the garage is cleaned before then.

Pull out phone. Open Motion app. Click ‘New Task’. Click title bar. Type ‘Clean garage before Motte comes in to town.’ Scroll down. Select a due date. Realize that your wife put it on the calendar, but you don’t know exactly when he’s coming. Close Motion. Open calendar. Swipe forward to “Dad coming to town for a few days” event on the family calendar. Note the date. Close Calendar. Open Motion. Pick a due date before he comes to town. Mark Medium priority. Click Save.

“I need to clean out my garage before Motte comes into town.”


“Ok, it looks like your father in law is coming into town on the 29th, so I’ve created a high priority task in Motion with a deadline on Saturday the 27th and titled it “Clean out the garage.”

This is not about time savings. People don’t mind spending time on something worthwhile. This is about 

 Elegance is simply speaking your mind. Telling a competent executive assistant your need and desire in that moment, in the same way that you communicate the rest of your needs and desires: language.

There are plenty of instances in which language is a terrible UX, and visuals are needed. For instance, reviewing your calendar for the upcoming week, or checking the status of a project by glancing at a Kanban board. So, there will absolutely be a long tail of use cases for “software UI.” However, I think these are a minority of the cases.

An even smaller minority are the times when a UI cannot be “dreamed up” on the fly by your AI for you to interact with. Perhaps it is more convenient for you to tab through a “New Lead” form in Salesforce than to try to explain it in a sentence, or click a toggle switch instead of saying “I want to turn off email notifications.” But who makes that toggle switch? Who decided whether or not it was worth building? Who decided the specific order of those Salesforce forms, or which ones even needed to be included in this specific experience?

90% of UI for SaaS applications is either unnecessary (can be replaced with sentences), or can be generated on the fly.

Interacting with software should start with the user’s intent, described in English. That’s how we already conceive of the action in our head. 

 can vary so much from user to user, but software companies have to design something that works for everyone. The experience of software needs to be brought closer to the user. You should have an AI that is advocating for you all the time, and constantly trying to accomplish your goals, as expressed by you. When it is inelegant to express something in English, your AI bot can craft UX for you which makes it easier to express what you want, or express data which needs to be added to an application.

Another thing that sucks about software is that it can only be applied to the really big problems in your life. We’ve all heard and personally had “app ideas.” The problem with nearly all of these ideas is that they’re solving a problem that is too small. How is it possible that an idea is good, but too small? Because human attention is finite. In order to solve a customer’s problem, you have to advertise to them, then get them to buy, then get them to remember to use your product to solve their problem. All of this costs the user’s attention, which is being competed over by a million other things.

There are already an unbelievable amount of software products in the world, and with the price of software production dropping, there are about to be even more. There are micro-SaaS solutions popping up for every conceivable inefficiency. In fact, there are probably 

 unique SaaS solutions for 100 aspects of your life that you don’t know about, because it’s too much effort to try to apply their benefits and keep track of them all. 

People are not sophisticated enough buyers to take advantage of the emerging benefits of ubiquitous and cheap software.

 There are many things which it is simply obvious for software to do. Previously, if software could improve the lives of some number of people, entrepreneurs could start a company around that and deliver that value. Now, there are a million small ways to deliver value; but there’s not enough attention surface area.

AI has the potential to give attention to more aspects of your life and match you with more beneficial services. 

AI is a more sophisticated buyer, and user, of software.

 It can take your general intent, and on the fly purchase 4 micro-software solutions to your problem and use them seamlessly.

To enable this, we need a marketplace for AI. We need somewhere that your AI can go and purchase the solutions to your problems. That might simply mean that it needs an adapter to Salesforce so that it can add a lead for you. Or it might mean that it actually needs to search for and buy a physical product.

I believe the future of software looks like this: you have a “chat bot” which is unique to you. It understands your life and your context. It searches the marketplace of products, APIs, and even other AIs to accomplish your goals. It pays for those services using money that you give it. It ties these products together and uses them elegantly to accomplish your goals and answer your questions.

 for human <> software interaction. This is first and foremost a language interface, but traditional UX can also be created on the fly when needed.

 for your personal AI to find the products and services that you need to get what you want.

 Developers of products, APIs, and AIs need to be able to list what they are capable of.

 The AIs that are tied to humans need to be able to find those services on the fly when they are tasked with something.

 AIs need to be able to pay the APIs and AIs for the services they use.

 AIs need to be able to rate products and services on how helpful they were in accomplishing their respective users’ goal.

 Once a service is chosen, your AI needs to actually be able to talk to the service, or to other AIs.

 Your AI needs a “scratch space” in which to tie together services, do computation on results, and cache data that does not belong in any particular service.

At first, I think the marketplace will primarily contain “adapters” to existing software products. However, I think eventually there will be many products created that are native to the platform. These will start out being the small one-or-two-time use case products. Remember that AI is a sophisticated buyer. It will suddenly make sense to create that micro-SaaS you were thinking about starting on the weekend; a million AIs can understand why what you just created is valuable, and pay you for 10 minutes worth of usage, a few times a year or less.

However, I think eventually we will see larger and larger softwares take place in this paradigm, with less and less proprietary UX. The fact is, everyone knows how a CRM should function. There is nothing secret about Salesforce. We know the schema. We know how the data should be transformed. Salesforce wins because users know its name and now how to operate its UI to their advantage.

But what happens when you don’t care about the software, you’re just telling your assistant that you want to capture a lead? Does it matter where the lead is stored? Owning the UX means you own everything.

Why Pretty Much All Software Sucks

When I was in high school and looking at the world of business from the outside, I always wondered why companies needed money or 

 at all. Wasn't the whole point of a company to generate cash, not use it? There always seemed to be something broken about a company that was a couple years old and raising a round of financing. Imagine how surprised I was when I came to understand the public markets, and found that even large, established companies were raising and using money all the time.

Of course, there are plenty examples of pure bootstrapping, where a founder simply grows the company organically, initially with his own money and then on the cashflow of the business. But as it turns out, using other people's money is something that most companies do most of the time. Why is that? Why don't they just use their own? The reason has to do with the function of capital in a society. And so in order to answer this, we need basic philosophy of capital.

There are many kinds of capital. I would roughly define capital as "existing value." You can have relational capital, which is the present amount of good will that another person has toward you. You can have social capital, the present amount of trust and good will that many people have toward you. And most relevantly to this discussion, financial capital is the present amount of good will that nearly everyone in a society has toward you, on the plane of 

Types of capital can be exchanged for each other, with some skill. Social capital can be exchanged for financial capital by means of carefully attaching yourself to a product. This is what sports players and actors are doing when they perform in advertisements, or when people who got famous in a viral YouTube video try to sell an NFT. In the other direction, financial capital can be also exchanged for social capital, like when businessmen build community parks or contribute to a new performing arts center. 

The most difficult and risky direction is between relational and financial

. Some people try to buy relational capital with a specific person using money, for instance buying expensive gifts for a girl so that she will agree to go on a date with you, or throwing a lavish party to impress specific attendees. This rarely ever works in the way the purchaser intended. The opposite direction can be even more perilous: trying to cash out relational capital into financial capital. Perhaps the most common example of this is multi-level marketing schemes. The reason that these are broadly disliked is that, while the participants often make money, they nearly always are simply exchanging relational capital they had with their friends for financial capital. Their friend might buy from them, but there is a high chance that their respect for that person decreased by the exact amount of the purchase.

There is almost always a lot of risk and friction involved in exchanging capital between types. The community that receives the park from the billionaire might feel that their affections are being purchased. You probably thought less of the famous actor after seeing them in the toilet paper commercial. And even after spending on a nicer car, only a few people actually respected you more than before.

That's not to say that you should never make these type of exchanges, but there is nearly always loss in the exchange itself. The only reason you should do it is if you have a very good reason to need to hold the other kind, and the gain you stand to using the other form is many times more than the friction loss.

 capital within the same type in order to gain more. For example, inviting someone over to dinner is a way to stake relational capital. By reaching out to someone, you experience a bit of friction loss, and more importantly, you put your relational capital with that person (respect, good will, etc.) at risk. If the dinner goes well, you could end up having more capital with your guest. If it goes poorly, you will have taken a step backward.

Social media is generally the realm of staking 

 capital. Every time you post, you are risking your reputation with some subset of people who will see your post. If your post is good, you can gain good will from many people at once. If your post is bad, you can lose a lot of it at once. In my observation, most people when they are new to social media start off by losing a lot of the social capital that they already had from the offline world, since it's really hard to look good in photos, videos, and words. Over time, as they figure out how to do social media well, they eventually tend to gain the capital back, and probably a good bit more than they had at the beginning. 

After this point, skill and effort will determine whether or not they continue to accumulate capital or will stagnate.

 Given enough time, everyone on social media will settle into a certain stable amount of capital which is consistent with their character and effort. Some posts might gain capital, but there will be an equal number which deplete it.

Perhaps the clearest type of staking is financial. If you want to make gains on the stock market, you must stake some of the money that you have in order to gain it. If you want to make money in real estate, you have to put some money down on a mortgage or a downpayment in order to realize the appreciating value of the home.

I think you can see by now that absolutely everyone engages in the 

 of capitalism in some way. The reasons that people do it are varied: greed and vanity are certainly very common. But other motivations are present as well: the joy of building something valuable, the feeling of doing something you're good at on a higher and higher level, and finally altruism, since you can give capital away to other people after you have accumulated it.

If you want something to be different than the way it is now, you must stake something.

 If you want more friends, you have to stake your existing relational and social capital, for example by asking to be introduced to new people, or by potentially looking green at a new type of social event. If you want more influence in society, you must use the influence you have now to climb to a higher position, or use it to kickstart a project with great social value.

This is the most fundamental reason that those who have capital stake it: to create change. Most often, the change they desire is to own more of the same type of capital. 

 One dollar now for ten dollars later. Ten million dollars now for a better climate later. Ten thousand dollars now for an educated child later.

The nature of staking is lossy by default. You no longer control the capital, and it is at risk. In order for staking to be successful, it has to be applied to an 

. This is the context which is carried in the phrase, "to capitalize on something." Perhaps you've heard a friend say something like, "hey, you're really good at talking! You should 

 and get into sales!" To capitalize on an advantage means to ascertain some kind of inefficiency, find an agent who has a unique ability to solve the inefficiency, and to then to double down on him so that he has the specific resources to solve it.

 of what is inefficient and who has the advantage is the true value behind successful capitalizing. (Any gaggle of MBAs can start and run a commodity business which already has generally known procedures. The returns on this will be marginal.) 

 First, they can see inefficiencies in the market. Second, they can identify actors who have a unique advantage which will help them solve the inefficiency. Finally, they put their capital in the hands of those actors.

 Those who do this successfully do it by giving capital to motivated agents who have an advantage in solving a market inefficiency. But what does this look like from the side of the agents? Why do agents need capital?

The reason is that market inefficiencies or "problems" have a certain size which is correlated with the nature of the market, 

. This sounds obvious enough, but really let it slosh around in your head a bit. I see young potential founders make this mistake all the time. They have a vision for owning a company which looks a certain way, and that company vision is somewhat static. (50 employees, a building downtown, golfing on the weekends). They also have a variety of startup ideas, and they imagine this theoretical company solving the issue du jour.

The problem is that different market inefficiencies have different natural sizes in the market. If you believe that groceries should get into people's hands differently, that is a multi-billion dollar problem. If you think that your small town could use a gas station on 3rd and Main, that is of course much smaller.

 within a market to solve a given inefficiency do not necessarily have the correct size. In fact it's very unlikely that they currently do, or that their liquidity lifecycle is aligned properly. The two things are completely uncorrelated.

That is why outside capital is needed. Capital helps you to 

 of a bet by right-sizing your resources to solve it. 

Using other people's money allows you to efficiently convert a potential advantage into an actual one, at the correct size.

In the next section, I'll discuss how each stage of company raises money, who they raise it from, and why.

This is perhaps the least understood capital transaction. There seems to be a mindset among young founders-to-be that capitalists are ready to write a check to anyone with a good enough idea. While it is true that the quality of your idea matters, who you and your co-founders 

 matters just as much. Idea-phase capital must underwrite two risks at once: that the inefficiency or problem is real, and that the agents in question have the ability to execute. Remember, capitalists are people who stake 

 by right-sizing an existing agent with a particular advantage. 

Whereas other companies have advantages pertaining to the 

, in the idea phase all advantages must lie in the 

That's why you'll often see high-profile or repeat founders raise large institutional money for their ideas, whereas less high profile or less experienced founders must work in smaller tranches.

In this phase, capitalists want to see a big problem and a competent team. The exact solution to the problem matters much less; it will likely change in later phases. It is easy to spot green founders in this phase. They talk about their solution as if it is an incredibly valuable thing, worth investing in, rather than talking about how big the problem is and why they are the right person to go and solve it.

What does the idea-phase capitalist look like? I would break them into two categories. The first is 

, high net worth individuals who have some specific knowledge about the industry. The risks that the idea-phase capitalist needs to underwrite are team competence and size of problem, and so angels will often have some technical experience that allows them to evaluate the team, or industry experience which helps them size up the problem.

It's very difficult for institutions to benefit from idea-phase capital over time due to 

. An individual angel might be able to take the highly dynamic set of variables involved in a founding team and an idea, but institutions which are naturally limited by the principal-agent problem have a very difficult time making repeated good judgements in this space.

institutions tend to provide capital to the idea phase in the form of 

 An accelerator does two things for the capitalist. First, it brings the founding team close enough to analyze over a longer term by junior people. Second, all accelerators push founders to describe their startup in terms of pre-defined metrics. These two things help increase the legibility of your startup for the capitalist, and investment will generally happen after a few weeks or months of this "reading" process. In exchange for this, accelerators offer resources that might help you push your startup forward, and a clear path to funding.

The startup phase is all about product-market fit. The capitalist at this phase is still underwriting the risks from the idea phase (can the team execute and is the problem real), but at this stage the solution also comes into play (does the team have a product which fixes the problem?). Product-market fit at this stage is usually described in terms of "early traction." Some examples of this are pre-orders or other forms of early revenue, email list subscriptions, and customer LOIs. Capitalists capitalize advantages. The advantage being analyzed in this phase begins to be the positioning of the actual company and product in the market, rather than simply team competence.

Unlike the idea phase, the startup phase (which I would loosely describe as pre-seed through series A) has a deep bench of institutional resources available to qualified startups. The best VCs of our generation such as Sequoia have made all of their greatest wins in this phase. It is early enough that the company still cannot be comprehensively described in terms of metrics, which means that well-networked or high-reputation players stand to benefit. Capitalists in this space stay deeply embedded in the startup ecosystem, constantly consuming and understanding the nuances that will allow them to correctly identify the winners in team, problem, and product.

A startup-phase company is laser focused on creating a product which solves the problem, selling it to a small set of initial customers, and then observing what happens next.

 A company can enter this phase once they are capable of talking to customers every day and writing code every day. If you cannot do both of these things every day for any reason, you are still in the idea phase.

To graduate from the startup phase, you will have shown a product that is currently being sold to customers at a high margin, which is measurably fixing their problems or providing much more value than it costs. Ideally, you are also showing a strong path to acquiring these customers at a much lower cost than the amount of money you make from them over time. If you can demonstrate these things, your company is in the growth phase.

The growth phase is the sexy phase. This is where everyone is having the most fun and bragging about their success on social media. Rounds in this phase, series A or B to IPO, range from the tens of millions to hundreds of millions.

The purpose of the growth phase is to pour gas on a scalable business model, with a proven product, fixing a big problem.

 It is not inaccurate to say that both of the previous phases were really just attempts to mix all the right ingredients together to reach this phase. Now is when the 

 that was initially conceived of in the idea phase is clearly demonstrable and actionable to outside capital. There will still be rapid product iteration during growth, but the main focus will be scaling the product that has been developed in startup phase to as many customers as possible without degrading the acquisition economics.

Interestingly, in the growth phase, the attention switches back to the "team" again as it was in the idea phase. The question is now, can this team create a 

 that triples in size every year for many years in a row? Can they hire good executives? Can they build processes that scale? Can they keep a level head with hundreds of millions in the bank?

Capitalists at this phase are almost entirely institutional. Growth-stage companies can be described by their metrics (MRR growth, CAC:LTV, churn, etc) with increasing accuracy, making it much easier for institutions to capitalize your company with larger amounts of money. Institutional money tends to crowd into the growth phase, which can create some perverse incentives to 

than actually makes sense. Remember, growth is sexy. Growth stage companies just seem to go up and to the right. LPs desperately want to have their money in tech startups, but the startup phase takes too much work and is 

 difficult to apply large amounts of money while beating the indexes. On the other hand, mature, post-IPO companies are obsessed over by hedge funds and retail investors, creating ruthless efficiency.

The growth stage is the landing pad for the hopes and dreams of every LP who just saw the YoY return of Apple stock since 1980. 

For this reason, the growth stage is constantly in danger of being overheated.

 Founders in this phase must be very careful not to develop a preference overhang by listening to the hype of everyone around them.

In the final phase, companies have an ongoing cycle of liquidity that can last for hundreds of years in some cases. The cycle goes like this:

Current management spots a new market or opportunity.

Management raises capital, either by selling its stock or raising debt secured against it.

If all goes well, they now have excess cashflow to payback the loan, or can sell stock at a now higher price to pay it off. If the market recognizes this positive movement, the move will pay off in terms of a higher share price. If the market lags in this understanding, the company can capitalize on that, too, by executing a stock buyback and then waiting.

However, if the opportunity goes poorly, the company now has debt with no excess cashflow to pay it, or has sold shares to raise money, with no new added share value to match that dilution.

Shareholders are either happy with or angry with steps 4 or 5 respectively. That anger translates into board elections which translates into officer elections, and the liquidity cycle starts over again with new management.

The story of a public company is the story of its managers' battles with changing market conditions and successful or failed execution of new opportunities. 

Companies that consistently win see an increase in the overall value of each share over time.

 (As an aside, don't be fooled by an increase in market cap. If a company raises money by issuing new shares and selling them, the market cap will increase by the amount of money added to the company, without any change in the share price yet. The move will only have been a success if the money that they receive from doing this capitalizes an 

 such that the company is able to realize more value than they gave away by selling stock.)

You might be wondering why I'm assuming that all "mature" companies are public. There are many reasons to want to stay private, but they mostly have to do with the owner's specific goals and desires for the company. All else being equal, being a public company massively increases your options, and generally lowers your costs, when it comes to liquidity. Private companies must continually work with banks or other large financiers to manage their liquidity, rather than simply disclosing their progress to the public markets and trading in their own securities. Even if a public company wants to use debt, the process is simpler since they can secure the debt on stock rather than arguing to the bank about why they should underwrite the next phase of growth.

So who is the capitalist in this final phase of a company? Well, you are. The market is. Everyone who has a 401(k), a pension, or of course owns a stock or piece of a mutual fund is a public investor. At this phase, 

companies have their greatest level of legibility, which means that inefficiencies in liquidity and pricing are solved with ruthless accuracy and speed

. The two main categories are so-called "smart money" buyers and retail investors. Smart-money buyers are institutions full of smart analysts which trade the public markets with deep professional analysis that helps them make good bets over time. Their 

 is to provide accurate pricing to each company, and they are rewarded in proportion to how correct they are versus anyone else. A hedge fund that has consistently great returns was simply consistently correct about the true value of the companies that they traded on. This pricing allowed other retail investors to make better decisions about where to put their money, and the institutions are rewarded for that service.

The other group of capitalists in this phase are "retail investors." These are individuals who buy stocks in small quantities in order to see gains in the value of their assets or cashflow over time. Their function in the market is to analyze the growth claims of individual companies, the value of their products, and to contribute to or detract from the price of those companies. Once again, they are rewarded in proportion to their judgement about the correlation between the company and its share price into the future.

The mature phase, characterized by the public markets, experiences the most efficient and liquid capitalization. 

This has some profound benefits and dangers to all involved.

 The benefits to the capitalists are the fact that all companies in the public market are required by the SEC to disclose information material to the position and quality of the company. This means that public companies are at peak legibility, which allows high level investment decisions to be made. Investors at this level can easily invest in "trends," can de-risk or "hedge" other investments, and can rapidly shift capital around as their judgement dictates. The downside is that all inefficiencies are highly competed over, and you must be very smart or very lucky to make serious gains.

For companies, having your stock available on the public market means that you can right-size your company and capitalize on advantages at a much lower cost. Stock-secured corporate bonds have some of the lowest interest rates after treasury bills. And selling your stock and then buying it back can, with some skill, create enormous value with no downside at all.

On the other hand, there are two major disadvantages to being public. The first is that you have an increased regulatory burden imposed on you by the SEC and other trade organizations. This can slow your organization down, and it can also limit your ability to make moves in secret. The other is the principal-agent problem. Company officers are often compensated with company stock. This is good in theory; the leader of the company should make it more valuable over time. But the issue is that it can incentivize officers to do things in the short term that increase share value, but harm the company in the long term. Unfortunately, some form of this problem will always exist for mature companies, except perhaps for closely held multigenerational family-owned companies.

One of my greatest learnings over the last two years is that 

. I don't mean this in the ideological sense, I mean it in the functional sense. Capital is 

. Here are some examples of things which are not capital:

 for value, not value that has already been created in the past and can now be used. Many of my business failures have come from thinking in terms of supporting businesses as a capitalist, but I don't have enough capital to do that effectively yet. Instead, my value comes from being an 

. I can use capital as a tool while finding efficiencies, building advantages, and then scaling. But skills and my time spent applying them are not themselves capital. I showed earlier how capitalists are people who stake 

This has several interesting implications. The first is that I must be far more picky than a capitalist is about what I involve myself with. A great capitalist can deploy capital into hundreds of companies. But if the value you bring to the table are things like skills, experience, or wisdom, there is an extremely limited subset of opportunities you can transact with. Don't be confused by the people who are in a different category. While people whose main function is applying capital jump from company to company on a weekly or daily basis, you should focus on only a couple of things where your talents are leveraged maximally and allow the capital to yield over time.

An interesting question that occurs to me every so often is whether I will ever be a capitalist, or if I will always want to build things myself. I can't imagine not being the builder, the agent. But I suppose many young people have felt similarly, and then, proportionally to the count of gray hairs, find that they prefer to operate at a more abstract level. I suppose that time will tell! Either way, I think I'll enjoy the ride.

Who gives companies money, how do they use it, and why?

An Overview of Modern Capitalism

This post is an attempt to categorize AI as an economic actor alongside more recognizable market forces. My goal is to contextualize what is otherwise a highly confusing and dynamic shift, without underrating any of the true novelty of what we're experiencing. Some of my thinking on this has developed more rapidly since the popularization of LLMs, mostly due to the fact that there are far more examples and use cases to consider now. But most of it I've been developing for a few years.

Inflation is when the supply of money increases relative to the amount of valuable things ("goods" is a sufficient term) in a market. The most common example of this is when governments print money. But the opposite can happen as well: the goods side of the equation can increase while the amount of money stays the same.

, and it happens whenever technology advances. When the cotton gin was invented, it suddenly became cheaper than ever before to buy cotton-based materials. Everyone had the same amount of money, but that amount went further (in the cotton market at least) than ever before. TVs are a modern example of this. It used to cost thousands of dollars to purchase a flat screen TV, but now anyone can have one for a few hundred dollars. Imagine for a moment if everything decreased in cost as fast as TVs have over the last 15 years. Economically speaking, this would present as a few thousand percent currency 

I believe AI could end up being the most rapid deflationary agent in history.

 In my own industry, software development, I estimate that it's now 30% cheaper to get a given software product created than it was 2 years ago. There are currently models being trained to create realistic CGI for movies, which could reduce the cost of making a movie by 90%. According to the National Private Truck Counsel, driver-related costs make up 45% of trucking costs. Without human drivers, or more likely with a remote human driver monitoring many trucks at once with the help of self-driving AI, the majority of that cost could disappear.

To get an intuitive model of why these falling costs are equivalent to deflation, imagine having $1,000 in a "rainy day" piggy bank sitting on your desk over the course of a couple of years. At the beginning of the period, your thousand could purchase a certain amount of goods and services from the market. 50 boxes could be shipped to you across the country, or you could purchase a certain number of newly built pages on your website. Now imagine that you push off spending your piggy bank for a couple of years, and by the time you get around to it, the AI improvements have set in. What can that same $1,000 do now? You can now have 10x more boxes shipped to you, or add 20 pages to your site instead of 2.

Here's another way to think about it on the macro scale. Let's say you had to sell everything in the world tomorrow in the world's largest garage sale. At the end of the sale, you will hold all of the world's money (naturally, since you hold all of the world's goods). Knowing this, you try to price every good and service in the world. You have to assign a relative value to each thing, and then proportionally spread the global amount of dollars across each good and service.

Now, imagine the same garage sale happening in 10 years following a period of AI development, in which more goods are created at a lower cost. Assuming the same number of dollars, what will have happened to the value of the average item? It must necessarily decrease relative to each dollar, since there are more items to spread all the dollars across.

So AI is deflationary from an economic perspective. The interesting part is what deflation really means for all of us.

It's interesting to think about who technological deflation benefits and who it harms. Based on our previous example, you might be tempted to say that deflation benefits those who are already wealthy, since their wealth is suddenly more meaningful. And this 

 true about those who hold their wealth in 

. But there is an interesting side effect to tech deflation that adversely effects the holders of equities.

The first of these problems is something that I'm going to call "

," because I didn't go to business school and I don't know the real term for it. I would characterize margin collapse as the opposite of "margin compression." Margin compression commonly occurs in an inflationary, rising cost environment; it happens when the cost of delivering your services rises faster than you are able to raise your prices. By contrast, margin collapse, by my definition, occurs in falling cost environments, where the acceptable price for your product falls faster than you are able to reduce the costs of delivering it.

Technological innovation causes broad-based margin collapse for existing industries as competitors arise using new methods which have a lower cost basis.

 This is a lose-lose for equity holders in those industries. Even if you manage to beat back the competition by adopting new technology and lowering your prices, you must spend capital to modify your product or processes, and that expenditure is simply a loss. Many companies will fail to adopt the new tech at all, and slowly experience margin collapse until they can no longer do business.

This understood, let's go back to the question above: does technological deflation benefit the wealthy? 

 Some already-wealthy people may be smart enough to invest in companies which use the new process and experience gains in market share (especially if they ride the wave early), but for the most part, existing companies will lose market share and have to spend capital to keep up.

There is also the possibility that cash, gold, or equities in uncorrelated industries may benefit from AI deflation. But in the era of bad money, large holdings of gold or cash are quite rare. And I'm not sure that there 

 going to be such an industry as one uncorrelated with the development of AI.

In contrast to equity holders, people with smaller amounts of cash savings, cash income, or workers on the forefront of new technologies have the tailwind of deflation at their back. Their salaries will go farther, their savings will be able to buy more, and their early stage stock in the next big thing becomes more valuable as market share is captured.

(As an aside, everything said above can be reversed for inflation. During inflationary times, your salary and savings mean less every year, whereas large public companies are able to use practically free capital to expand indefinitely.)

One of the common fears I hear about the onset of AI in all facets of the economy is the complaint of centralization. The assumption is that since only the largest companies seem to be able to build and operate large AI models, the propagation of AI into every space will mean centralizing power for the owner of the models, or the political actor willing to seize them.

There is merit in considering how to prevent such massively centralized outcomes. But I think that the economics will do a better job of preventing this than anyone. Consider that 

all technological disruption starts out centralized and becomes increasingly decentralized over time.

 When Gutenberg first created the printing press, he had the only one in the world. If you wanted a book, you 

 to go to Johannes for it. However, this paradigm flipped on its head as the new discovery moved down the economic chain, and now there is practically an uncountable number of ways that you can send and receive written information (like this blog).

This chain of economic development holds true across all sectors. Here's how I like to formulate it:

Scientists -> Inventors -> Technologists -> Industrialists -> Commercializers

Or, as seen from the perspective of the consumer:

Discovery -> Invention -> Application -> Product -> Commodity

I believe that today, AI is somewhere between the invention and application phase of this progression. Because of this, it is still centralized, living partially in the labs of the inventors, and partially in the hands of the few technologists who understand how to apply it.

It is clear to me that the situation will not stay this way. AI tech is going to be decentralized very quickly as it moves down the economic chain, primarily by the industrialist. We will soon see industrial-scale AIs fine-tuned to every type of use case, and finally commercialized solutions at diminishing cost which abstract all of those different options into an AGI-like "intelligence".

AI as a base service is quickly becoming cheaper as compute cost drops. OpenAI doesn't have anything incredibly unique in their underlying methodology, so it's only a matter of time until more competent models can be grabbed off the shelf from any number of vendors. For this reason, I don't think of AI as a centralizing force in the long term. Instead, 

it seems to be shifting power away from large entities and toward individuals.

 Large technology organizations are in danger from small teams of people who can use Github Co-Pilot effectively. Hollywood studios are in danger of people who are prompting entire movie scenes into existence. Google is in danger from You.com and Bing. Rapid technological development is anti-entrenching; as we saw above, it erodes the cost structures that allow a single player to go unchallenged for years.

Every good or service that exists today is over-priced

. Don't believe me? Ask any 90 year old whether they wished they had spent more time earning money and buying things, or more time with their kids. At the asymptote of technology, every possible good or service has a value approaching zero when compared with the finitude of human life.

A more humorous example: there's a story going around right now about a woman who is angry because Tesla dropped their prices 25% just after she bought a brand new model. Her anger makes for a funny story, but she is right: the Tesla was overpriced when she bought it. What she and everyone else are missing is that the Tesla is 

When we buy a product, what are we actually buying? The communists would say that we're buying labor. We're not, but that's a post for another time. 

 If you disagree, I've got a lump of copper, a chunk of aluminum, and a handful of sand to sell you. I call it an iPhone. It's the same exact materials as the thing they'll hand you in the Apple store, but mine has a 20% discount. What, you won't pay $800 for that? iPhones are the physical form of thousands of man-years worth of wisdom about how to turn rocks and minerals into human productivity and entertainment. When we purchase an iPhone, we are purchasing wisdom. When we purchase a hammer from the hardware store instead of a stick and a lump of iron, we are purchasing the 

The fundamental reason that software has been so mind-bogglingly profitable, for both equity holders and workers, is because 

software allows you to rapidly embed wisdom into the physical world, and wisdom is the only thing in the world with value. 

AI advances this principle by an order of magnitude. Previously, hundreds of engineers were needed to embed a small piece of functional wisdom into the world. As of a few years ago, you could embed the same wisdom with only 10 engineers.

But even those 10 engineers know that a huge amount of their time is not applying the core wisdom of the business, rather it is spent following the motions of common sense. So much of company building is simply "embedding" what a company already knows. This is why so many ex-entrepreneurs turn into venture capitalists. It rewards them for being 

 (their core value to the world), but allows them to skip the years of simply trying to apply the insight that they had in a few moments.

The promise of AI is to radically decrease the amount of time that is spent on simply carrying out the mechanical actions that you already have in your mind. In the same way that software engineers today simply write code and don't think about server administration, engineers of the future will seek and embed wisdom, and not think so much about loops of code. (As an aside, I don't think that this will put 

 software engineers out of jobs. Good engineers already know that their value comes from 

deciding what to code, not the actual typing.

Since I was young I have always felt that 

 was important. With experience, I have been surprised to find that 

 is what really counts. I believe the underlying reason for this is that 

As technology develops, we march toward more and more direct application of the mind and wisdom to the world. Every year it becomes more a world in which the mind rules. In which the mind and soul gain power.

I am incredibly excited about the years in front of us. Remember Bill Gates' insight: we tend to overrate the 2 year impact of a breakthrough in technology, and underrate the 10 year impact of it. I'm seeking to build a philosophy of technology that makes sense of the latter outcomes. Here's to the next 10 years!

Why everything in the world is overpriced, and how AI could change that.

AI and the Value of Everything

Over Christmas break I put out a Facebook post which proposed the following thought experiment:

Let’s say you had a Starbucks in a mall on the day before Christmas. There are more customers than you can possibly serve. You’re gonna make a VERY nice Christmas bonus today.

How do you decide to run it? Do you:



 Keep the line as short as you can, take as many orders as possible, and then allow people to sit down on your tables and couches to wait for their orders, or...


Let the line grow as long as it can, only taking new orders once old orders have been fulfilled?

By the way, this is not a question about profit vs customer experience tradeoffs. The correct answer makes you more money *and* makes your customers happier. The wrong answer makes you less money today, makes your customers mad, and hurts your repeat revenue (negative NPS score for this cohort of customers).

The Starbucks I just went to chose A. Were they correct?

I was pleased to see people opining in the comments. Apparently I am not the only one using my Christmas break on fun inanities! However, I was also very surprised to see how strongly the majority opinion tended toward method A. Here are the results by my count:

Method A won the popular vote by a margin of 25 points! That's ok. I knew that I was going to be the contrarian on this one.

There are two knobs which can be turned: profit and customer happiness. (I take Starbucks to be sufficiently standardized that drink quality is constant).

I actually think that the revenue is just about the same for both A and B. This is because revenue is determined by the bottleneck, which as far as I can tell is the espresso machine. Assuming you have saturated demand all day, you can roughly estimate the revenue by the following: 

[Flow rate of bottle neck] x [Time open] x [Average ticket size]

Good restaurant flow probably means making lines longer, not shorter.

Why Pretty Much All Software Sucks

Thursday, July 27th, 2023

An Overview of Modern Capitalism

Sunday, February 19th, 2023

AI and the Value of Everything

Monday, February 13th, 2023

In Defense of Long Lines

Sunday, December 25th, 2022