It's a memoir by a computer nerd who bought into the startup myth, and then was very disillusioned very quickly.
I worked at a pre-IPO software startup. We IPOed, but at a low price. The investors did not like that price. In a bad business, we pivoted a better one; we found a small-but-solid income there. But investors on the board of directors were looking for high-risk, high-return; they weren't interestd in small-but-solid. So the board fired our CEO and we got a new CEO who promised to move the company in a new direction with unknown, possibly high-return potential… which didn't materialize. The company went out of business.
There was already a startup myth floating around: You just need to make it to the IPO; then you sell some stock and you're set for life. But I'd been through an IPO. I'd seen the company fall apart. I'd learned that an IPO wasn't the goal, wasn't a guarantee of future success, just a milestone.
There were booms and busts. There was a lot of investment in startups selling things on the web before many customers were actually set up on the web. New graduates started showing up wanting to work in tech who weren't enthusiastic computer programmers. They were excited to make money; back in my day, such folks would have gone off to work in finance or at Bain. Now these folks who equated financial success with real-life success… these folks were coming to work with me.
Among the booms and busts, the startup myth evolved. If you were in a startup, you just needed to convince a clever venture capitalist, titan of industry, that your startup was poised to change the world. Then your company would get funded. These wise venture capitalists would then guide your company's expansion; soon enough, you'd be making the world a better place on worldwide scale. It was risky: 9/10 startups went under! But as long as one company returned 10x, it was all worthwhile.
That sounded pretty good to me, a win-win story. Someone more VC-savvy than me (Rob Hayes, maybe?) said: hang on. Think about that. The VC is excited about your company because they think they can get a 10x return on investment. If you do better than 90% of startups and survive but you only manage a 2x return, those investors might think you're overlooking some important business. Remember when you worked at that company whose board fired the CEO so they could get a new CEO with an overall-stupid-but-with-high-potential-upside plan?
Anyhow, I reckon that I took a leisurely fifteen years to get disillusioned about the myth of the startup. My lessons were abstract; I lived with the dread of knowing that my employer could go under any month now. But I wasn't leading those companies; I didn't feel responsible for steering them towards sustainable business.
OTOH, this book's author is a computer nerd who, as a university student, bought into the startup myth, helped form a startup, helped lead that startup as CTO, then had the myth forcibly torn down around her over just a couple of years, feeling responsible when the myth failed to materialize. How did this happen?
As a grad student, she worked on inference engines. She wanted to make a living doing that. An inference engine applies rules to data. E.g., maybe you have a rule "All men are mortal." If you know that Socrates is a man, you apply the rule to conclude that Socrates is mortal. If you have a huge distributed database of two billion thingies, about a billion of whom are men, then an inference engine efficiently applies your rule to tag your billion mortals.
At Google, I worked with some scholarly folks who were eager to work with big data. Instead of working at a startup, they worked at a big company. They went to Google and other big internet companies because those companies had lots of data to process. Nobody spent years studying computer science to quickly determine that Socrates was mortal; to make things interesting, you want to look at a billion of Socrates' friends, too.
But according to the startup myth, boring risk-avoiders go to work for the already-big tech companies. The elite technical people start startups. So the author formed a startup to apply her inference engine knowledge. As a startup, she didn't have any data to process. She might have hoped that big companies would hand over big data sets with requests that she write software to process that data; but big companies weren't eager to just hand that data over.
So her startup scrabbled to find work. They paid for data from social networks, processed it to find insights about the users of those social networks.
They fell into doing work for marketing companies. Marketing companies want to understand people. If they have a rule like "People who enjoyed the Candy Crush game will also enjoy this new puzzle game" they want to apply that rule to figure out who to market this new puzzle game to. This can get into some shady privacy-violation stuff if you're not careful; like, say, if you stumbled into this business because you wanted to tinker with this really interesting piece of technology; you didn't get into this business because you're that eager to find out who's into Candy Crush. If you fulfill marketers' requests without double-checking those requests against privacy violations, you will violate privacy all over the place.
One day, the author looked over the Terms of Conditions for a social network her startup was getting data from. Thus, she found out her startup broke the rules. Her startup had been violating those rules for a while. This wasn't part of some evil Cambridge Analytica-like plan. It was just an easy problem to fall into if you assume surely there will be some safeguards in place, a warning message will pop up somewhere.
So, because she'd followed her startup dream, now she had the dread of keeping her company going though its business was built on using data in forbidden ways; she risked disappointing her investors; she risked unemploying her friends and colleagues…
The author was so discouraged that she went back to school to study anti-capitalism. (I didn't know you could study that‽) In theory, money measures something's value to society, because people pay for what they value. In theory, if you want to make the world a better place, you figure out how to make the most money and that means you're on the right track. In practice, money's a darned inaccurate measure; externalities add up. The startup myths center on the money: get that VC funding, get that IPO. But if you chase the myth, you can find yourself violating your own principles for no-good-reason.
Anyhow, this was an exciting read, sorta like how a horror movie is exciting when you want to yell, "No, don't go in there!"
