From The Wall Street Journal:
Spiders can spin intricate webs. Birds weave branches into cozy nests. Bees build hives with near-perfect hexagons. It should be easy for an advanced robot complete with lasers and artificial intelligence to lay a simple brick wall. But, as the journalist Jonathan Waldman chronicles in “SAM,” the quest for a bricklaying robot has been bumpier than the work of a mason with vertigo.
The tale follows Scott Peters, a 30-something engineer in western New York and co-founder of Construction Robotics, as he spends most of the past decade developing SAM (Semi-Automated Mason), prodded by the inspiration and funding of his architect father-in-law, Nate Podkaminer. We watch as the two create a company, hire engineers, experiment with marketing and finally stack walls, haltingly.
Several themes run through the book. First is the often-unsung adaptability of organic intelligence. Engineers have sought for decades to make devices that build with blocks. “Getting an inanimate machine to do what only hands and brains could was apparently some kind of universal geek fantasy,” Mr. Waldman writes. “And while it sounded like child’s play, it was phenomenally difficult. To put it in context: The first machine that successfully picked up small wooden blocks did so only eight years before humans landed on the moon.”
The minute adjustments a human makes when manipulating objects, especially in messy environments like construction sites, result from billions of years of evolution. We make it look easy, until you give instructions to a robot and watch it fumble around or freeze up when it gets a little dirt on its face. Yann LeCun, Facebook’s chief A.I. scientist, once told me, “I would declare victory if in my professional lifetime we could make machines that are as intelligent as a rat.”
Mr. Peters has laudable motivations. “By creating a bricklaying robot,” Mr. Waldman writes, “he aimed to eliminate lifting and bending and repetitive-motion injuries in humans; to improve the quality of walls; to finish jobs faster and safer and cheaper; and to ease project scheduling and estimation. Basically: to modernize the world’s second oldest and most primitive trade.”
. . . .
The robot’s development is an object lesson in debugging. The engineer who crafts the code for the arm has the six stages of debugging posted in the office: “1. That can’t happen. 2. That doesn’t happen on my machine. 3. That shouldn’t happen. 4. Why does that happen? 5. Oh, I see. 6. How did that ever work?” The team swaps out parts, rewrites code and reconfigures designs, sometimes in rain or under a blazing sun. Often the solution is an utterly familiar one: Switch the damn thing off and on again.
A second theme is that technological advancement requires debugging not only hardware and software but also humans. Some of that problem-solving is simple workflow optimization: loading bricks and mortar properly and promptly, keeping people out of the way. Some of it requires deeper psychological and sociological renovation. It’s hard to smooth relations with workers who don’t want to share a job with SAM in the first place. Some masons said its work was sub-par (but more colorfully). Some said it threatened to steal their jobs. And some just didn’t like doing things differently. “The construction industry, the engineers began to learn, was as slow to change as baseball,” Mr. Waldman writes. Similar lessons are emerging from the front lines of semiautomation in medicine, manufacturing and other fields.
. . . .
Mr. Waldman follows all the drama like a fly on a brick wall, richly reporting scenes and conversations, many on job sites where both circuitry and civility break down. The book is reminiscent of a reality-TV show about a scrappy startup, complete with backstory segments as we learn the pasts and personalities of each new hire. There are also a lot of digressions—the history of the bricklayers union, how much pinboys at bowling alleys were tipped, how literal sausages are made, Mr. Peters’s 16th-century ancestors, his high-school swim coach’s career as a famous-in-Japan professional wrestler.
Link to the rest at The Wall Street Journal (PG apologizes for the paywall, but hasn’t figured out a way around it.)
PG has worked with very bright inventors a couple of times during his business life and has found them to have a unique mindset and combination of talents and personality traits.
Of course, authors are a species of inventors, creating stories out of keystrokes.
Both authors and inventors rely upon the protection of intellectual property laws – copyright and patents – to permit them to control and exploit their creations in the way they deem best.