Books: The soul of a new machine

TL;DR: Great book about engineering corporations. If you work in Big Tech, you should read it.

⚠️ Warning: Spoilers

I enjoyed The soul of a new machine, by Tracy Kidder.

Though short, this historic account of the development of a 32-bit computer in the late seventies explores a large number of interesting topics about engineering corporations.

The book is very factual, a recounting of the design and implementation of this computer, sans preaching. You draw your own conclusions, mostly. The exceptions towards the end, where Kidder ventures opinions about such things as the role of computers in society, are few and short —and stand out so strongly against the rest. Kidder is almost recounting relevant opinions of others. They compliment the story well.

The story shows various forces in tension. In the management style of the Eagle team —the Eclipse Group, lead by Tom West—, the great directional freedom given to engineers, and the tremendous delivery pressure. Nearly everybody in the book appears both tremendously proud of being in the team, and exhausted from work nearly all the time.

Two divisions in different sites of the same company, compete directly with each other. The book follows the smallest of the two, West’s group, as they implement a machine they expect will become very significant. They are trying to keep their plans under-the-radar from VPs and execs until they get far enough to be impossible to be shut down. The corporate politics make the company appear somewhat dysfunctional. Towards the end, however, Kidder speculates that this competition across divisions might have been a deliberate decision by the CEO —and possibly even a reasonable strategy. However, the Eagle team getting disbanded post-launch makes this somewhat unlikely (why disband the winning team?).

The book gives a great accounting of many sources of intrinsic motivation: the elegance of a design or implementation; the desire for “revenge” against the “enemy” division —we’ll show them fools what we are capable off—; the great freedom and responsibility —many of the engineers, freshly out of school, were given tasks significantly larger than anything they had accomplished—; or the feeling of finding ourselves at a unique point in history and being given the chance to create something awe-inspiring. The desire to build beautiful, elegant machines, which transform the world for the better, is a tremendous motivator.

The book also serves as a great illustration of how a single engineer, West, can be instrumental to the creation of something so complex that no individual could create alone. It shows very technical persons having tremendous impact through abstaining from getting involved too closely in technical considerations, beyond few simply-stated high-level principles (e.g., De Castro’s “no mode bit”).

It’s also sobering to see up close that the sense of urgency we feel these days with the pace of development around AI —the tremendous pressure to work long hours and weekends, meet deadlines or all is lost— is nothing new: it was already clearly there, strongly shaping West’s team, half a century ago. West had to commit to impossible deadlines due to the organizational politics he navigated, and used them as a way to put pressure on his team.

Fascinating that it was all so ephemeral. The team of experts quickly put together produces great results and promptly gets disbanded. And, in the end, the company mostly fades away in an acquisition (for unrelated reasons).

Related