The Soul of Machines
Modernity is defined by the computer. Over the course of my lifetime, they have moved out of the shadows of government, education, and business to become completely implanted into our lives. Software shapes our experience in ways undreamt of by even the most prescient science fiction writers of the past, and the hardware that it runs on — the wires and chipsets — are the rails that power most of the modern economy.
Because of this, it’s easy to forget that all this infrastructure was decades in the making. Turing’s magnificent machine might have changed everything, but it took many more years and countless other minds to make the journey from his rolling mechanical bits to the Occulus Rift and the impending horrors of the metaverse.
The earliest computers were highly specialized, and saw their most immediate applications in warfare. One of the earliest was a machine known as the Torpedo Data Computer (TDC), designed to calculate firing solutions for submarines during World War 2. Only a couple ever saw service, but they represented a considerable improvement over earlier methods of calculation (done by slide-rule. Mind you — some really impressive things were done with a slide-rule). The were also mechanically complicated, bulky, and difficult to manufacture. In the early days, every computer could be considered a custom job.
The TDC was an analogue system, meaning that it relied on mechanical parts to store and manipulate information. Technically, some version of this has been with us for a long time, but the TDC served as a bridge into the digital age. The first digital computers, known as the Colossus series, were designed in Britain to break encryption cyphers during the war. They proved quite an adept at cracking German codes and contributed mightily to the allied victory in Europe. American engineers would soon unveil their own version: ENIAC, a programable calculator that the army used to plot artillery trajectories — just in time for the end of the war. ENIAC weighed 30 tons.
The first commercial computer, called the Ferranti, was only produced twice, and served principally as academic and business tools. The doors were opening though, and a new industry was born. The company that manufactured it (Ferranti, later Ferranti-Packard) would go on to build a number of successor models, but never in large numbers. These early computers (or computing complexes) paved the way for a revolution. Improvement built on improvement, and advances in core technologies made miniaturization possible. Complexes gave way to rooms gave way to cabinets gave way to desktop terminals gave way to laptops.
The race was on, and very quickly a host of companies had sprung up to capitalize on some really radical improvements to produce a growing diversity of commercial models. The story is well known: computers got smaller and more flexible. The competition was cutthroat, but the rewards (for the few) were great. Many of the companies that drove these developments are still with us, many more have been forgotten or folded into other companies. The history of computing is a compacted history of iteration — design, test, deploy, repeat. Engineers were often hobbyists first — the kinds of people who cobbled together circuits in their basements. This was the era that brought us Radio Shack (if you remember that place) — the first computer scientists often had to build their own devices simply because no market was as yet in place.
In 1981, writer, scholar, and journalist Tracy Kidder embedded herself with a company called Data General, a leading computer manufacturer. Data General was a pretty big deal back in the sixties and seventies, and produced a series of cabinet-sized “minicomputers” used in a variety of applications. Kidder’s resulting book The Soul of a New Machine, provides a fascinating look into the personalities, challenges, and organizational landscape of the early tech industry, many elements of which, for better or worse, remain embedded in the culture.
Most engineers are pretty passionate about their work. This is only follows — a kind of energy is needed to really master any science. Intuition is a cultivated talent and is only developed through constant experimentation. Consequently, the engineers at Data General are constantly citing endless working hours dedicated to unrealistic project schedules. And their managers demonstrate remarkable self-awareness of it. Most of the engineers, of course, won’t receive anything special for the work — stock options, kickbacks from unit sales — these things are reserved for the managers and salespeople. Instead, they convince themselves that the work is the point (“I’m not doing this for the money” was a common refrain among them — especially upon finding out how underpaid they tended to be). And management consistently shows a keen awareness of this fact. An awareness and a willingness to take advantage of it.
And perhaps this is the most revealing element of Kidders’ investigation: exploited passion is a basic piece of the operating culture in most companies I’ve worked for — especially startups. One popular (business) terrorist manual, Blitzscaling, is shot through with recommendations for tormenting employees into compliance by creating an environment built on instability and frustrated expectations. By keeping people insecure and overworked, you can help ensure that they remain loyal right up until the moment you don’t need them anymore. In an era of disposable labor and cheap capital, this is a blueprint for the quick build, fast sale. It also helps ensure that no company is actually responsible to it’s obligations. You don’t blitzscale to create something lasting — you do it to make a quick buck at other’s expense.
So Soul of a New Machine tells two stories. In the first, a group of talented (albeit overwhelmingly white and male) engineers and managers created one of the first 32-bit computers. But the other is a tale of cynicism and exploited youth — kids that don’t know better trusting to men that have no interest in their figuring it out. When you reflect on modern companies and their cultivated burnout culture (the great resignation has been a long time coming), it becomes painfully clear that the future of tech will need to find it’s way back into the hands of the makers. After all, employees that aren’t invested in the organization they work for are more likely to cut corners, look the other way, and otherwise be less responsible with things like private data.
Nevertheless, it is the creators that vest these machines with life. Software is a creative invention, requiring dozens of minds to design, code, and deploy. And creation is, at it’s best, an act of love. The lesson of Kidder’s book, and of the modern employment landscape, is that we must take better care of those that invest social products with their soul. We need to do better about honoring idealism — if nothing else to better look out for those that are looking out for us; the ones that will do the job well regardless of the money (ed. they should probably get more money too).
