In August, I cavalierly said that an AI couldn’t design a car if it hadn’t seen one before, alluding to Henry Ford’s apocryphal statement “If I asked people what they wanted, they’d say faster horses.”
I’m not backing down on any of this, but the history of technology is always richer than we imagine. Daimler and Benz get credit for the first automobile, but we forget that the “steam engine welded to a tricycle” was invented in 1769, more than a century earlier. Assembly lines apparently date back to the 12th century AD. The more history you unpack, the more interesting it becomes. That’s what I’d like to do: unpack it—and ask what would happen if inventors had access to AI.
Learn faster. Dig deeper. See further.
If Nicolas-Joseph Cugnot, who created a device for transporting artillery on roads by welding a steam engine to a giant tricycle, had an AI, what would it tell him? Would it suggest this combination? Maybe, but maybe not. He might have realized that this was a bad idea – after all, this proto-car could only go 2.25 miles per hour and only for 15 minutes at a time. Teams of horses would do a better job. But there was something about that idea—though it seems to have died—that stuck.
During the last years of the 19th century, Daimler and Benz made many innovations on the way to the first machine generally recognized as an automobile: the high-speed internal combustion engine, the four-stroke engine, the two-cylinder engine, the two-cylinder engine. rotary steering, differential and even a gearbox. Several of these innovations have appeared before. Planetary gears harken back to the Greek Antikythera mechanism; two-pivot steering (placing the knuckles on the wheels rather than turning the entire axle) came and went twice in the 19th century – rediscovered by Karl Benz in a trade magazine. The differential dates back to at least 1827, but probably appears in Antikythera. We can learn a lot from this: It’s easy to think in terms of individual innovations and innovators, but it’s rarely that simple. Early Daimler-Benz cars combined a lot of newer technology and redesigned a lot of older technology in a way that was not expected.
Could a hypothetical AI have helped with these inventions? It might be able to resurrect two-turn steering from “winter steering”. It is something that has been done before and could be done again. But that would require Daimler and Benz to get the challenge right. Could AI have invented a primitive gear, given that watchmakers knew about planetary gears? Again, the exhortation would probably be the hardest part, as it is now. But the important question was not “How do I build a better driving system?” but “What do I need to make a practical car?” And they would have to come up with this challenge without the words “automobile”, “horseless carriage” or their German equivalents, because these words were still being created.
Now let’s look ahead two decades, to the Model T and to Henry Ford’s famous quote “If I asked people what they wanted, they would say faster horses” (whether he actually said it or not): What is he asking? And what does that mean? In Ford’s time, automobiles as such already existed. Some of them still looked like horse-drawn buggies with motors attached; others looked recognizably like modern cars. They were faster than horses. So Ford didn’t invent the automobile or the faster horse – but we all know that.
What did he invent that people didn’t know they wanted? The first Daimler-Benz automobile (still in modified buggy format) predated the Model T by 23 years; his price was $1000. That’s a lot of money for 1885. The Model T appeared in 1908; it cost roughly $850, and its competitors were significantly more expensive ($2,000 to $3,000). And when Ford’s assembly line went into production a few years later (1913), he was able to lower the price even further, eventually bringing it down to $260 by 1925. That’s the answer. What people wanted, what they didn’t know they wanted, was a car they could afford. Automobiles were firmly established as luxury goods. People may have known they wanted one, but didn’t know they could apply for one. Little did they know it could be affordable.
That’s really what Henry Ford invented: affordability. Not the assembly line that first appeared in the early 12th century, when the Venetian Arsenal built ships by lining them up in a canal and moving them downstream as each stage of their production was completed. Not even the automobile assembly line Olds used (and patented) in 1901. Ford’s innovation produced affordable cars on a scale previously unimaginable. In 1913, when Ford’s assembly line began production, the time to produce one Model T dropped from 13 hours to about 90 minutes. But what is important is not the time spent building one car; it is the rate at which they could be produced. A Model T could roll off the assembly line every three minutes. That’s the scale. Ford’s “any color as long as it’s black” did not reflect a need to limit options or cut costs. Black dried faster than any other color, so it helped optimize assembly line speed and maximize scale.
Of course, the assembly line wasn’t the only innovation: Replacement parts for the Model T were readily available, and the car could be repaired using tools that most people already had at the time. The engine and other important sub-assemblies were significantly simplified and more reliable than those of competitors. Materials were also better: the Model T used vanadium steel, which was quite exotic in the early 20th century.
However, I was careful not to attribute any of these innovations to Ford. It deserves credit for the biggest of images: affordability and scale. As Charles Sorenson, one of Ford’s assistant managers, said: “Henry Ford is generally regarded as the father of mass production. He wasn’t. He was his sponsor.”1 Ford deserves credit for understanding what people really wanted and coming up with a solution to the problem. He deserves credit for realizing that the problems were cost and scale and that they could be solved with an assembly line. He deserves credit for putting together the teams that did all the engineering for the assembly line and the cars themselves.
So now is the time to ask: If artificial intelligence existed in the years before 1913, when the assembly line was designed (and before 1908, when the Model T was designed), could it answer Ford’s hypothetical question, what do people want? The answer must be “no”. I’m sure Ford engineers could have used modern artificial intelligence tremendously in designing parts, designing the process, and optimizing the work flow along the line. Most of the technologies had already been invented and some were well known. “How can I improve the carburetor design?” is a question that AI could easily answer.
But the big question – What do people really want? – it is not. I don’t believe that an AI could look at the American public and say, “People want affordable cars, and that will require making cars at a scale and at a cost that is not currently conceivable.” The language model is built on all the text that can be scraped, and its output in many ways represents statistical averaging. I’d be willing to bet that a 1900 era language model would have access to a lot of information about horse maintenance: care, disease, diet, performance. There would be a lot of information about trains and trams, which are often powered by horses. Some information about cars would be found, especially in high-end publications. And I imagine there would be some “wish I could afford one” feeling among the growing middle class (especially if we allow hypothetical blogs to go with our hypothetical AI). However, if a hypothetical AI were asked what people want for personal transportation, the answer would be horses. Generative AI predicts the most likely response, not the most innovative, visionary, or remarkable. It’s amazing what it can do – but we also have to recognize its limits.
What does innovation mean? It certainly involves combining existing ideas in unlikely ways. This certainly includes resurrecting good ideas that never made it into the mainstream. But the most important innovations either do not follow this pattern or complement it. They involve taking a step back and looking at the problem from a broader perspective: looking at transportation and realizing that people don’t need better horses, but affordable cars in bulk. Ford may have done it. Steve Jobs did it—both when he founded Apple and when he resurrected it. Generative AI can’t, at least not yet.
