Mike Smitka
Autos are today a hi-tech industry; more on that in subsequent posts. The same was true circa 1900.
A quick window on the contemporary mindset lies in the pages of Scientific American. The late 19th century was the world of Thomas Edison (incandescent lighting was first used in the 1880s), of Dunlop and rubber (1889) and bicycles (1885 for the first chain), of more powerful (steam) locomotives, of battleships and bridges. Searching the titles shows only scattered mentions of automobiles. During the first half of the 1890s, the hot topics were developments railroads, followed closely by marine transport, and then steel and electricity. By the end of the decade, topics electrical dominated, with railroads close behind, followed by bicycles and then ships. Electricity remained the hot topic at the start of the 1900s, but now automobiles came in second, followed closely by ships. Railroads were fading as an area of interest for technologists of the era; bicycles were a mature technology. The last four years of the decade saw electrical developments as the most common, followed by railroads, automobiles and ships. Throughout the period 1890-1908 advances in transportation were presumed a major interest of readers.
Detroit was the Silicon Valley of the 1905 era
So did the venture capitalists of the day. Ford's first commercial attempt at making automobiles (1899) failed within 2 years, and he soon left a second venture in the hands of unhappy investors to found today's Ford Motor Company in 1903. That startup began with $28,000 in seed money. That was a tidy sum – a $2.50 workday and a six-day workweek meant a decent job paid $750 a year – but not out of reach for wealthy investors.
Ford was not alone in this. While the modern automobile originated in Germany, and many of the key technical advances were then made (and first commercialized) in France. Europe though was fragmented, and remained so through 2003, when the end of the Block Exemption finally allowed a fully integrated market across the continent. The U.S. was a bigger market, unimpeded by tariffs and in sheer numbers a larger middle class, the China of its day. Entrepreneurship was everywhere – new towns needed their retailers, farmers needed to sell their produce to people who could store and ship it, and provide financing; reapers and other new-fangled agricultural implements needed sales, financing, repairs. Dynamism is both provides the opportunities and is an outcome. Cars were just one avenue.
As with Silicon Valley, the industry was populated by serial entrepreneurs and networked individuals. For the US, Thomas Klepper [2002] analyzed data on 725 ventures for which sufficient records survived to confirm that they sold at least one vehicle. (Others have subsequently expanded that list.) About 120 were manufacturers of other products that tried their hand at cars; another 145 were spinoffs from existing car companies. Finally 108 were headed by entrepreneurs with prior industry experience. In total, just over half of all ventures were part of this broad start-up community.
Early entrepreneurs had to battle with the physical layout of cars, which gradually evolved away from being mere horseless carriages, and with what would provide the motive force – steam, electricity, or internal combustion engines. It took until 1903 for the industry to focus on the latter. Who would buy cars? The early market was partly for well-heeled enthusiasts, spurred on by racing. It took a while for them to become practical, but by 1905 there was a reasonable chance that someone would be able to both get where they wanted to go, and return. Dreamers could easily point out the limitations of horses: costly to keep up, a public nuisance because they were inefficient in eating [almost a pound out for a pound in], limited in distance, and needing a wagon to convey more than a single rider. Automobiles were touted as the way of the future.
Initially these ventures were scattered around the US, with only 1 of the first 69 firms during 1895-1900 located in Detroit. But by 1913 there were 41. In a later study, Klepper [2007] traced the process of spin-offs and the movement of experienced managers from existing firms to new firms. Detroit however had a diverse and large manufacturing base, including rolling stock for railroads, cast iron products such as stoves, shipbuilding for the Great Lakes, and other diverse industries. It was centrally located, and large enough to provide sufficient labor. Other industries, such as the (horsed) carriage makers of Flint (home to Billy Durant), were nearby, and there was a bounteous supply of wood for making car bodies. With the success of Packard and Olds, this base proved larger than in New York City, St. Louis, Chicago, Rochester (NY) or Indianapolis, which were the other five early centers. (This is from Cabral et al. [2013] who revisited the issue to focus on determinants of firm survival.) These all had a broad base, but through chance a couple Detroit-based firms did better, and spin-offs (such as Chrysler) were more numerous and more likely to survive the initial start-up period. The semiconductor industry saw entrants such as Texas Instruments outside of the Bay area, but that developed into Silicon Valley. The same thing happened with autos and Detroit.
Nor is venture capital new. The public, or at least the stock market, believed that cars would be the wave of the future before the rise of the "killer app" of Ford's Model. Entrepreneurs could aim for an IPO at an early stage; Billy Durant was able to scrape together a bunch of such ventures in 1908 to create General Motors, which effectively failed in 1910 and again in 1920 only to rescued by creditors. Investors wanted into the new hot thing – which included the formation of other "general" "American" and "trust" agglomerations such as General Electric (1896), American Sugar (1891) and American Tobacco (1890). Durant and his General Motors were part of a boom of such firms in the years leading up to WWI that found a ready market for their shares, even if there was not much of a market for their products.
In sum, from the perspective of technological innovation, venture capital finance and a geography-based entrepreneurial start-ups culture, Detroit was the Silicon Valley of its era.
| Year | electrical | rail | ships | automotive | steel | engines | bicycles | wood | chemistry | TOTAL |
| 1890-1894 | 132 | 252 | 207 | 53 | 132 | 89 | 37 | 35 | 25 | 962 |
| 1895-1899 | 246 | 219 | 135 | 51 | 85 | 74 | 162 | 25 | 11 | 1008 |
| 1900-1904 | 382 | 216 | 129 | 259 | 92 | 70 | 22 | 20 | 23 | 1213 |
| 1905-1908 | 197 | 185 | 111 | 139 | 57 | 51 | 11 | 15 | 16 | 782 |
| Total | 957 | 872 | 582 | 502 | 366 | 284 | 232 | 95 | 75 | 3965 |
Citations:
Cabral, Luis M.B. M., Zhu Wang, and Daniel Yi Xu. “Competitors, Complementors, Parents and Places: Explaining Regional Agglomeration In The U.S. Auto Industry.” National Bureau of Economic Research, 2013. 6114.
Klepper, Steven. “Disagreements, Spinoffs, and the Evolution of Detroit as the Capital of the U.S. Automobile Industry.” Management Science 53, no. 4 (April 2007): 616–31.
–––––. “The Capabilities of New Firms and the Evolution of the US Automobile Industry.” Industrial and Corporate Change 11, no. 4 (2002): 645–66.
Why is this so great? Chips this powerful can span multiple displays in the vehicle. You could have an infotainment system and a digital instrument cluster running on a single i.MX8. Because you don’t have to worry about virtualizing a single GPU (which is quite the challenge), you can carve up the chip’s graphics and processing power to isolate the infotainment system from the cluster. This, in turn, minimizes your scope of certification. Achieving ISO 26262 for a cluster is daunting enough; achieving it for a complex infotainment system as well is off the scale. 
