Widgets and Generators

by Mike Shea on 27 December 2013

While we seem to be fascinated by the idea that physical robots are stealing our jobs, software will increase the gap between corporate profits and individuals trying to earn a living . We built educations and careers around building widgets — physical products or services paid for by the hour of labor. Now and in the future, sustainable livings may only come from generators — creative works copied and distributed at no cost that generates income with every transaction. Unfortunately, few will build generators solvent enough to sustain a living. Who can say what happens to the rest of us.

On Jobs, Corporate Profits, Robots, and Software

I'm fascinated by the topics of robots taking over jobs and the gap between corporate profits and individual employment in the US. Here are a bunch of interesting articles on the topic:

It's certainly not clear that all of these things are tied together, but together, they paint an interesting and potentially bleak future for knowledge workers.

The idea that robots will take over jobs of manual labor, software stands to disrupt knowledge work far faster. Building robots still requires a fair bit of resources and labor. Developing software only requires a handful of programmers to build software able to reach a billion people.

This is why MP3s destroyed CDs. Even though the music industry had streamlined the process of producing and shipping CDs to almost zero, it still didn't come close to the efficiency of distributing MP3s over Napster and, eventually, peer-to-peer services.


For the sake of this article, "widgets" represents products or services paid for by the unit or by hour of labor. A coffee at Starbucks, a wreath hanger from Amazon, or tires for your car are all widgets. There's a whole manufacturing process to make them and there are a million ways to make it cheaper, most of which include taking away jobs. A lot of salary-based knowledge work is also in the production of widgets. Knowledge workers often get paid to produce a thing, a digital or physical paper, metadata in a database, or some other bit of information. Knowledge workers are also paid for by the hour to continually serve the need for which they are paid.

A widget is a good or service for which the creator only gets paid once.


A generator is a creation that can be infinitely copied and generates income for each copy. Books, movies, music, licensed artwork, TV shows, software patents, iPhone apps, and hats for Team Fortress characters are all generators. Creators make them once, can create and distribute copies for nothing, and can generate revenue for each one distributed.

Ebooks are a great example of a generator. An author writes a book, hosts it on Amazon, and can sell as many copies as there are people willing to buy it without a single other human being needing to get involved. There's almost no cost to replicate and distribute it. You can write a book and sell a billion copies without having to construct factories or develop entire new shipping platforms.

Constructing, selling, and shipping physical books still requires an infrastructure full of people. As Planet Money's T-shirt venture showed, getting the physical product to the door costs more than the material costs. In many cases, moving a thing around costs more than constructing a thing in the first place.

As we watch robots take over the widget-making industry, it seems obvious that humans should get into the generator-making business. While there have been some forays into computer programs turning into generator generators, I think we're a ways away from a computer program automatically constructing a hit youtube series. Maybe not that far, however.

What About the Other 150 Million?

There's no way that everyone who used to generate widgets for the man is going to turn into Hugh Howley and write a hit ebook like Wool. The business of building generators is full of a tiny amount of big hit makers who never have to work again and millions of people who make much less than a living on it. Like every successful venture, generators still need an audience. Digital distribution can result in 10 times more profit than physical distribution, but that 10x multiplier isn't useful if the base number of sales is zero. Few people can run a Kickstarter backed with enough money to make a living for a year, much less continue to make a living for the rest of one's life.

Becoming independently solovent on personal generator production is as likely as winning the lottery.

What does the country do with 150 million people who can't live on profits from generators and were right-sized out of a job now replaced with computer software? I think that's going to be the big question of the next century.