Artificial Scarcity

Chase Raz

Chase Raz

Artificial Scarcity

Bison with its young in a field.

We talk a lot about the Artificial Intelligence (AI) that is being created thanks to our impressive technologies such as Machine Learning, but we never seem to talk about some of the other–more negative–artificial creations that technological achievements have brought about.

To fully understand this, we need to dip into economics for a moment.

Technology is very good at reducing variable costs. However, creating new technology products actually drives up fixed costs. This is why video games aren’t free… they take thousands, or even millions, of human hours up-front to create. Fixed costs, fortunately, are mostly distributed via economies of scale; and, interestingly enough, are reduced over time by leveraging the macroeconomic advantages of other innovations, but that’s an explanation for another day.

Technology is so good at reducing variable costs that they can nearly approach zero, especially for digital goods.

Imagine that you write a book.

In the ancient past, you’d have to find a scribe to copy your book word for word in order to make any sale past the first (the copy you wrote). In Ancient Egypt, successful families used to spend their entire life savings to pay a priest for a copy of the Book of the Dead which contained the negative omissions (e.g. “I have not stolen.”) to state in the afterlife to secure favor. It was an afterlife-insurance policy in a way, and it was expensive, especially considering it was often buried with the patriarch of a family. To put it colloquially, priests made mad bank copying and personalizing each scroll.

In the more recent past, you’d have found a publisher or printer to make your book using expensive machinery that the average person couldn’t afford. Technology drove up fixed costs by replacing one human scribe with machinery that required many unskilled laborers, and at least one skilled laborer, to produce. Great! We know we can offset fixed costs through methods alluded to previously.

Today, if you wrote a book, you could either print a copy yourself at home (another macroeconomic benefit), send your book to an on-demand printer like in yesteryear, or create a digital copy.

So, how much does that digital copy cost? Especially now that technological devices have proliferated (yes, macroeconomics again) and virtually every potential reader of your book has access to a display device?

Welcome to the present year!

The cost of a digital good, at least of any reasonable size, is a fraction of a penny for the storage space to hold the original; another fraction of a penny to store a duplicate copy for the purchaser, which is either offset to the buyer or is actually eliminated via centralized gated access like a subscription; and a fraction of a penny for transmission costs for delivering the copy or operating the subscription… either way.

Sometimes, these are fractions of a literal penny and not just a turn-of-phrase. In some cases, the three fractions don’t even equal a full actual penny. Still in others, like for video games, we may need to adjust to nickels or dimes.

Have all you want! This is Valhalla, and the feast never stops.

Let’s forget about your great book and now think about your favorite online service. This could be your web hosting through GoDaddy, your image creation through Canva, or your social media management through Buffer. Take your pick or come up with your own example if you prefer.

What type of Artificial Scarcity exists? You pay $15 per month for some digital good (again: hosting; design; or scheduling, respectively) that economically costs a fraction of a penny to deliver on-demand and in real time. Why can you only host one site; only get so many premium templates; or manage a certain number of social accounts?

GoDaddy makes you pay for a more robust account to host more than one site, and even hides certain features behind this elevated higher-paid gate. Canva allows only so many premium image or template uses. Buffer only lets you manage so many accounts and provides fewer analytical services. You know how entitlement tiers work.

And I’m not unfairly picking on these services… I’m just talking about them because you’ve likely heard of them. Almost every virtual good is in the business of pricing upon artificial scarcity these days. It’s not just GoDaddy, Canva, and Buffer by any stretch of the imagination.

You’re forced to pay an extremely high margin on a virtual good, but this is acceptable to many as the fixed cost of development is being addressed. But then, we have to pay substantially more to get a service limit or feature increase?

The fixed costs have already been dealt with by the markup on the base plan and the only additional variable costs are also minuscule (digital storage, data transmission, etc.). Certainly the bump up to the next plan should be a marginal increase, right? Instead, what do you often find? A multiple of the base plan’s price for any increased service entitlements.

This is a classic “Linear Programming” model (that’s math, not software development). With Linear Programming, something is being optimized, meaning minimized or maximized within a set of constraints using linear modeling. In this case, the current trend is to maximize by profit to the near exclusion of everything else, such as value, utility, adoption, benefit, etc.

So, what gives?

Humans and our economic wants are preventing the economy from being an agent unto itself in an unencumbered way. This isn’t necessarily bad… it just is. We’re accustomed to capitalism driving up prices through inflation as growth is required to survive. A fight against commodification and a need to stay balanced across high-tech and traditional industries lingers. Humans, at the end of the day, need to eat and have a place to sleep. We don’t like altering the status of markets, even when the inputs have fundamentally changed.

Ever wonder how so many people seem to be creating independent incomes and living off residuals? Yeah, it’s all related… we simply don’t need everyone’s labor. But, that’s a tangent and I’ll circle back to it in another post in the future.

Unfortunately, the costs of the human necessities (food, water, shelter, clothing, etc.) were set solely on traditional industries because they are the traditional industries. We don’t seem to be ready and willing to acknowledge that, or to truly adapt to a fully high-tech world. For proof of this, just look at the housing prices and the homelessness rates in two of the United States’ biggest tech areas: San Francisco and Seattle.

Houses and apartments can be constructed–using advances in technology–in a way that’s cheaper, faster, safer, and more luxurious than ever. We can build structures in a way that optimizes for a maximum population density while still providing ample personal space and comfort for the senses, but there I go with suggesting optimizations outside of profit again. We can even work remotely in many industries as a way to diffuse geographic supply and demand pressures. I mean, hell, we can practically almost print building rather than construct them at all.

Still, the cost of housing and real estate continue to increase in yet another bubble which is destined to literally destroy livelihoods during the next economic downturn… because there will always be a next economic downturn.

Getting to a Conclusion

You’ve likely picked up on my point by now and realize that book publishing and housing costs are just examples. The logic supporting these examples applies to almost every traditional industry prior to the digital revolution. Do healthcare, insurance, or traditional telecom services come to mind for you?

I’ll certainly be discussing this topic many times in the future, but for now it’s good enough to conclude with a prompt for you to think about. Something concrete and tangible. Here it is:

What would you pay for a house if you knew that it was built faster than ever, stronger than ever, by fewer people than ever? What if division of labor were offset and the work involved were similar to all work: interface with a digital tool which in-turn interfaces with other digital and mechanical tools? Sure the tools are expensive, economies of scale and macroeconomic benefits could offset them.

Would you still be okay spending a quarter-million dollars for a house? This, by the way is roughly the average home price in the United States.

A US$250,000 home represents four years of solid, non-stop, full-time work for the average U.S. household of 2.53 people and generally assuming between one and two adult workers.

The interest paid on such a home over a thirty year standard mortgage is an additional $300,000, or almost five additional years of work.

All things considered, the expenditure for a new average home is approximately $550,000. Think about that: an average of two adults, each with average incomes, working full time for 9 years each just to pay for a residence… not even the stuff to go in it, the expenses to upkeep that employment, or even food or water. That’s 18 full time human labor years just to buy a financed home. The financing is already worth more than the home, and that should tell you something about the future of work?

But again, this isn’t about houses, is it?

Under the same understanding, is it really sensible to be spending $15 per month to host one personal website? Business, sure! But personal?

What if you knew the rate used to be about $5 per month, even when the required technologies were more expensive? …and you could host as many sites as you could cram in! By the way, the monthly retail cost to acquire the computing power necessary for 100 of these sites today costs less than a cup of plain coffee at Starbucks.

This, ladies and gentlemen, is Artificial Scarcity (AS), and I’ll be addressing it more in the future, both positively and negatively.