Followers are likely expecting this to be my yearly newsletter on recent advancements in Artificial Intelligence.
The problem is, nothing unexpected happened in the world of machine learning last year. It’s been a boring year for technology. But it’s been a truly groundbreaking year for business in general.
Since you’re giving me a slice of your attention this day, let’s focus on real changes, and talk economics.
Here are quick-links to each section of this newsletter:
This starts out as a bit of a rambling walk, but I promise, it all relates together by the end.
If you follow only one reference link from this newsletter, please make it this one:
A 15-minute video isn’t your thing? Let me summarize it with a screenshot:
The gist here is that all the major US passenger airlines have valuations less than their own frequent-flyer programs. A lot less. Basically, airlines cover all their actual operating costs by selling frequent-flyer points.
That is, each airline is effectively minting their own currency, called points, which are redeemable only for that company’s services. Selling points is where the real money is. Flying passengers around is actually a loss-leader that helps them sell more points.
Said another way: airlines make money by literally making money.
Compared to this, exchanging services for dollars is such a dated business model that it has negative value by comparison.
I wanted to open this newsletter with airlines because they’re a great non-digital example of the modern economy. But minting new types of money doesn’t require billion-dollar physical investments to take part. If we allow ourselves to think digital, crypto-currencies let any of us play in this game.
I’m assuming we all know what crypto-currencies are: money, vaguely de-centralized.
Bitcoin is the old darling of crypto-currency. Since inception in 2009, the value of a bitcoin has gone from just pennies to 50,000 US dollars or more.
The technology behind crypto-currency is open to anyone though. For a while after Bitcoin proved the idea, any smart team of people with a lot of social connections kept trying to invent their own currency. For example, Litecoin, Ripple, and let’s not forget Dogecoin (pronounced “doggy-coin”). Dogecoin began as pure satire, yet Elon Musk pumps this currency regularly.
Ethereum was our most recent year’s darling. The Ethereum crypto-coin has been around a while, but the value went absolutely vertical this last year because the NFT gold-rush is carried predominantly on the Ethereum network.
We’ll talk about NFTs in the next section.
Fundamentally, none of these crypto-currencies are much different from the others. In all cases, people could theoretically use their home computer to burn power and turn it into digital currency, and other people would pay them units of that currency in order to record trades of that currency in a distributed database.
In practice, the people who dominate each currency network are not using home computers, but entire warehouses full of computers, sometimes built right on top of the gas wells that power their electric generators. (reference)
Also in practice, for the longest time there was nothing you could actually buy with your crypto-currencies, except other currencies. That’s started to change a little, especially because of NFTs, but predominately the whole concept is still one of pure currency speculation.
But it’s become big business nevertheless. Just as with airline points, it’s far easier to make massive gobs of money by directly minting money than it is by providing real goods or services.
Indeed, there’s a brain-drain going on in the tech industry, where engineers and executives are jumping from established companies to hop on the crypto bandwagon (reference)
Global supply chains for actual goods have been impacted as well. Are you, as I am, still trying to get a new Playstation? Well, crypto-miners are in line ahead of us. There’s more profit for microchip makers to sell to crypto-miners than to the makers of commodity goods. It was reported that of all the GPUs manufactured in the world during 2021, about ⅕th of them went to Ethereum mining alone. (reference)
Also, just as in the traditional stock market, the crypto-currency markets are now being dominated by flash-traders, who front-run the transactions of common people. (reference)
Big business indeed.
But crypto-currencies have been around for years. Why have they suddenly started to matter?
To answer that, we have to understand NFTs.
NFTs are brand new, so let’s spend some time trying to define what they actually are.
A logical first place to start is the acronym:
These words may perfectly document the concept, but they carry no reason we should care. Instead, proponents of NFTs like to talk about the problems that these things are supposed to solve:
However, I’d argue that there is a different set of problems that NFTs actually solve:
Let’s dive into details on each of these, starting with the acronym.
Here, token means crypto-coin. That is, an NFT is a single coin on some crypto-currency network, usually Ethereum.
The Non-Fungible part means that this specific coin is unique, and no other coin on the network can act as a substitute.
That is, when you buy an NFT, you’re buying a very specific crypto-coin that can not be mixed up with anybody else’s crypto-coin. Accordingly, the value of your coin is unique to your coin as well, and has no bearing on anyone else’s coins.
The biggest reason I hear for why someone wants to buy NFTs is because in a digital world, buying something is usually indistinguishable from renting.
Imagine you buy a movie, say from Disney. The actual digital bits that make up the movie do not live on your smart TV, they continue to live on some server in the cloud. But bits can’t just be made available to everybody. Therefore, each and every time you watch your movie, your smart TV has to politely ask Disney for permission to access those bits. If Disney (not always known for being fair with their business partners #DisneyMustPay) ever loses those records, you will no longer be able to play the movie you supposedly own.
If your movie purchase was an NFT, then hypothetically this scenario is solved, because the record of your purchase is stored on the crypto-currency block chain, which is supposedly indestructible.
The obvious short-sight here is that the NFT is just the record of sale. The actual movie bits still live in the cloud on Disney’s servers, and they could still remove those at any time.
Said another way, owning an NFT might make it slightly easier to find your purchase receipt, but Disney still controls an army of high-priced lawyers, and it’s pretty much up to them if they want to honor that receipt forever or not.
In other words: a record of sale carries no enforcement.
NFTs caught on first as an art market. Here, they’re trying to solve the problem of exclusivity.
Say you have a favorite artist, and you would like to support them by buying one of their works. Maybe you even commission something special. They deliver, and you’re delighted. Then later, you walk through the local art fair and see copies of your special unique work being sold, maybe even from multiple vendors!
On inspection, this is the same scenario as before: supposedly an NFT would solve the problem by being a permanent record of the sale plus all the conditions of that sale. But in practice, to write it out once again: a record of sale carries no enforcement.
This has recently become a very serious problem – not on the distribution side, but on the source side. It is common these days for bots to scrape major art outlets such as DeviantArt and steal thousands of works of art at a time to mint into NFTs. The artists are not involved, and receive nothing.
Further, the major NFT art site – OpenSea – has so far been minimally responsive to the issue. For instance, if an artist has had dozens of their works stolen at once, OpenSea is requiring an individual takedown case be opened for each separate item. A lot of Personally Identifying Information (PII) of the artist is required in each challenge. Plus, the word on the street is they are sharing this PII with the maintainers of the bot-nets who stole the art originally. (reference)
In other words, if you’re an artist, reporting art-theft is the same as self-doxxing to cyber criminals.
Tip for artists: OpenSea uses Google as back-end storage for all their jpegs, and some artists have been having more success petitioning Google directly for copyright takedowns.
The art industry has been deeply scarred. Some artists have totally closed their online profiles and preemptively removed all their works from every gallery site as a defense against this theft. This, of course, significantly damages their legitimate business. (reference)
I personally would like to see OpenSea go poof and disappear inside a class-action settlement of some kind. But whatever the outcome, to say that NFTs have disrupted the art world is a total understatement.
Next, let’s look at what problems NFTs are actually solving.
All that above is live right now on the digital streets. But what’s being talked about in the digital boardroom is scarcity.
NFTs are like trading cards. Supposedly, you buy and sell NFTs with your friends, trading pieces of art in much the same way people of various ages have traded baseball cards, MtG cards, and Pokémon cards.
But why stop at flat pieces of art?
What about trading whole custom characters, for playing inside a massively-multiplayer video game? What if those characters could transfer between different games, much like Wreck-It Ralph walking between game systems in the movies?
Or what about buying limited-edition upgrades to your new, real-world, electric car? What if those upgrades could follow you, and be available to you, even if you were riding in your friend’s car that day? Would that help you bond with your friends?
The term “Web3” is being used to describe these scenarios. Web3 companies are the new dot-com boom. That is, every week we see startups getting funded at massive valuations. (reference)
Scarcity has been a problem for tech companies for a long time. This is why software companies spent years converting their revenue models from selling boxes of software into selling software-as-a-service. But with NFTs, companies can sell individual units again, because there’s built-in scarcity. They need not even think in terms of whole units of software, they can do limited-edition runs of features inside that software.
This is the first deep allure of NFTs: they bring back scarcity in a digital world. (reference)
The second, deeper allure, is in how companies get funded at all.
When a company needs a big influx of money, it either takes out a loan, or as a last resort, it sells off shares of ownership. But selling shares of ownership is a highly-regulated process, overseen by powerful government commissions.
Funding your company by selling NFTs is not regulated, by anyone.
NFTs can be created in bulk – non-fungibility aside. As such, instead of issuing 1000 shares of stock, you can issue 1000 NFTs. You needn’t even give them all the same rights and privileges. As well, our legal system has no precedent to draw on when sitting cases about crypto-currency contracts.
To quote another author’s fine words on the topic: “Suddenly you have an asset class that you can buy from your portfolio companies that looks like a security, swims like a security, and quacks like a security, but is not regulated as a security. In fact it’s not regulated at all.” (Stephen Diehl)
We’re going to swing from talking about crypto-currency to talking about infrastructure. But as a bridge topic, let’s touch on information security, commonly called InfoSec.
The internet is a dirty place. I follow a number of InfoSec people on social, and there’s days I wonder how anything keeps running at all.
Never in the last year was this more clear than about two weeks before Christmas, when a critical vulnerability came to light called Log4j (pronounced: “log-forge”). This vulnerability hit deep. Companies from Apple to Microsoft were affected by it. A cyber-attacker could find a vulnerable system, and basically run any code they wanted as a full administrator.
When a hacker compromises a system, the next things they usually do are:
Ultimately, this doesn’t shut down business. Company resources are unsuspectingly being used to make crypto-money for some hacking group, but aside from a higher power bill, business can generally proceed.
If a hacking group is serious, and they can get deep enough access to critical systems, they do a ransomware attack. That’s where they encrypt all the compromised company’s data, then demand a crypto-currency ransom before giving over the decryption key.
Most companies find it simpler to pay the ransom than it is to rebuild their internal systems.
Ransomware is a ballooning problem. Ransomware-as-a-service exists – hacking groups who will execute a ransomware attack against a company of your choosing in exchange for Bitcoin. (reference)
This has been such a problem that business insurance agencies are beginning to require proof that companies are boosting their InfoSec defenses before they renew insurance contracts. (reference)
The overall topic makes a great segue into discussing infrastructure, for two reasons:
First, a lot of physical infrastructure has been impacted by this.
On the large end, a ransomware attack shut down gasoline delivery across the entire southeast US for six days last year. (reference)
On the small end, the electricity co-op in even the small agricultural town where I live had their billing systems taken offline for almost two months because of a ransomware attack.
Second, remember that Log4j vulnerability mentioned above, which left systems open to breach world-wide?
The actual vulnerability was inside a small library of open-source Java code maintained by about 5 unpaid volunteers. This just derives from the onion-like nature of much of our digital existence: outer layers depend on inner layers, and keep depending on inner layers, for as many layers deep as you look. Many of those inner layers do not get a lot of attention. But if one of them cracks, it can cause unexpected and massive costs to a company, or even our power grid.
If this isn’t critical infrastructure, what is?
My favorite economist is Jeremy Rifkin.
I first found his work early in the Covid pandemic. I was watching news feeds full of stories about our labor-force being out of work. And yet… the stock market was still going up.
It’s as if revenue is no longer tied to production, I hypothesized. Then I went looking for smart people who were talking about anything like this.
Jeremy Rifkin has spent years talking about this, among other things, to major world governments. He’s consulted for the US government, the European Union, and China.
What I learned from listening to his lectures is that digital goods have near-zero marginal cost, and this basically reinvents economics.
Rifkin further hypothesizes that the trend of distributing processes is going to edge out of digital space and into manufacturing, and then suddenly even physical goods will have near-zero marginal cost, moving us officially into a post-scarcity world.
In more recent writings, he calls this the Sharing Economy.
The recent work I’m citing here is a report he prepared this last year for the US Senate, called AMERICA 3.0 - THE RESILIENT SOCIETY.
His document is a detailed vision and plan for a massive investment in US infrastructure. If you want the highlights of his plan, they’re on page 7.
My reaction is that I started out reading the 242-page document and thinking it was boring. By page 13, I was getting mad.
To be clear, his document reads like a very detailed plan to build a new power grid. He also describes how to build a communication grid and a mobility & logistics grid in parallel with the effort. I’m fully confident these would be fantastic investments for our country.
I’m mad because his vision is flawed.
My complaints are as follows:
The America 3.0 plan imagines the Sharing Economy is going to come to dominate, and zero-marginal-cost economics will happen in all sectors. The problem I see here is that NFTs, and the entire Web3 concept, is geared toward manufacturing artificial scarcity. Given the recent massive business shifts towards scarcity, I’m highly skeptical that the Sharing Economy is as certain as the plan suggests.
For example, in the previous section I posed that information security is a large part of critical infrastructure. Even the America 3.0 plan tips to this, and has a part discussing Operational Security. The problem is, that part is about ½ page long and contains neither details nor even a strategy.
In fairness, the America 3.0 plan does include discussion of Social Infrastructure, to some degree, but that leads to…
Quoting the plan: “Social infrastructure includes broad investments in education, public health, affordable social housing, environmental protection, and other community services, all essential to assuring upward mobility.” (page 28)
Also: “Without universal access to broadband, the youth in disadvantaged communities will be unable to secure adequate careers in the America 3.0 infrastructure transition.” (page 30)
(Both emphasis are mine.)
In both cases, I’m reading that Rifkin cares deeply about giving our youth a chance to build a career.
My problem is that careers are not a realistic goal for vast swaths of our people.
We currently have five generations in the job market at once. (reference)
You can not convince me the elderly folk I see working as greeters at Wal-Mart are interested in building a long-term career.
On top of this, for just about the majority of everyone, most jobs are not worth being careers.
Here where I live, in a booming small town of Colorado:
Does Rifkin seriously believe education is the answer? Like somehow everybody in the whole country, from teenagers to grandmothers, can be educated out of having to do labor, and into some viable career that pays better decades down the road?
This doesn’t even address people who are disabled, nor the people who’s job is to stay at home and care for them. What upward mobility are these people supposed to seek?
I pose the America 3.0 infrastructure plan is:
I think my favorite economist has lost his way.
Our government passed a large infrastructure bill this last autumn. It was called the Infrastructure Investment and Jobs Act, or more commonly, the Bipartisan Infrastructure Framework (BIF).
Here’s the breakdown of what our elected leaders define as infrastructure:
Listed out, this is slightly more than half the money invested into transportation – roads, bridges, rail, and public transit. The remainder goes to electricity, broadband, water, and “resilience”. (reference)
This all sounds lovely.
The most meaningful criticism I’ve heard about it is that the bill included about 25 billion in subsidies to help green-wash the fossil fuel industry. (reference)
Overall, I guess it’s good, as far as it went. That is, this is just physical infrastructure, and not one bit of it is social infrastructure.
Social infrastructure was supposed to be addressed in parallel, through a second bill.
As I write this, the 2021 Build Back Better Framework (BBB) has officially failed. This is sad, because here’s what’s in the bill:
It’s worth telling that, on a line-item basis, ⅔ to ¾ of Americans support each of these things. (reference)
There’s been talk of resuming the effort in 2022, but my understanding of the political proceedures involved suggests that it’s a complicated issue. It has something to do with the filibuster, and how the only way to make a bill that can’t be filibustered out of existence is to call it a budget reconciliation. Yet there’s only two chances per year to do that, and since one of last year’s chances was already wasted on this effort, the powers that be are likely to be really scared about burning another opportunity for the same issue. (reference)
Be all that as it may, the takeaway is that the majority of our elected leaders have indicating that, to them, the definition of infrastructure does not include direct support to individual people.
The question I’ve been leading to this whole time is this: what counts as infrastructure?
Is infrastructure just roads, bridges, electricity, and water?
Or does it include the people who drive on those roads, as they run deliveries or staff our hospitals?
Are airlines infrastructure?
Are airlines still infrastructure, even if the economic value of an airline is the customer-base that they have locked into rebate programs, rather than the actual service they provide?
Do the 4 or 5 people who volunteer their time to maintain some obscure open-source software project – that just happens to be depended upon by major organizations world-wide – count as critical infrastructure?
Are insurance companies infrastructure? What if the insurance companies are the only agents in our country currently boosting our InfoSec “resilience?”
Are crypto-currency networks critical infrastructure? Will they become so, as the very concept of property ownership skews towards Web3?
Are the news networks infrastructure? Or the social media that we discuss that news on?
We all understand that our very democracy is under attack. But who you believe is doing the attacking depends on which channels you receive your news through.
We all know vaccines save lives. Yet here in my small town’s hospital, 120+ professional health-care workers have refused to get their Covid vaccinations.
In debates like voting fraud and vaccine efficacy, it seems very reasonable answers exist to “what is true,” and every challenge can be debunked, if we just aggregate the answers from a majority of the experts involved. Yet we have no universally-trusted network to reconcile truth and distribute the results. Instead, our economic systems and social media algorithms favor stirring up discourse, because that makes more money. (reference)
Can we really find no better metric to weigh conflicting truths than what makes us more money, faster?
How can we make any progress on big problems, from social infrastructure to climate change, when truth itself is bought and sold like a commodity?
I seriously doubt that turning news articles into NFTs will solve anything, but I’m absolutely sure we’re going to see it attempted, probably within the next year. (reference)