Longreads + Open Thread

Moon, Welch, Robinhood, Heterodoxy, Bots, Liquidity, Trade, Google

Byrne Hobart Byrne Hobart

Longreads

SPONSORED

You're on the free list for The Diff. This week, paying subscribers read about how endless ambient entertainment changes the market for low-skill labor ($), why some AI companies are gigs, not institutions ($), and how to think about the fundamenatls of the VIX ($). Upgrade today for full access.

Upgrade Today

Books

In the Plex: How Google Thinks, Works, and Shapes Our Lives: it's sometimes instructive to read business profiles a long time after the fact, to see what kinds of outcomes were visible before, what kinds were surprising, and what little details turn out to be significant. This book tells the story of Google's rise from grad school project to massive company, mostly with a focus on the early days.

Part of the fun is that early on, the story is one hack after another, at different levels of abstraction: Google was cadging computers from wherever it could, and assembling rickety-looking distributed systems that could collectively handle more than special-purpose hardware. But this also meant that they were inventing a lot of the science of reliability as they went along, and sometimes would run into problems like being unable to finish their crawler without the whole program crashing and needing to start over. (They clearly fixed this.) Other signs of Google's early compute constraints, no doubt formative for the company culture, included doing a deal where they powered search for Netscape and didn't have enough hardware to support both that and their direct-to-consumer business (they turned off the latter until they could get more machines).

There are some unbelievable stories of Google's early days and early culture. This one, for example, I just have trouble believing:

The need for some oversight became clear as early as 2000, when Brin sent employees an email announcing that Google had a new valuation (meaning the estimate of its market price had gone up) and would soon reprice its employee stock options—from 25 cents to $4.01. Some people didn’t realize that $4.01 was a reference to the calendar and frantically tried to buy up all the shares that they were entitled to before the price went up. They dug into savings and borrowed from their families. Google eventually had to make people whole.

If you take the abstract beats of the story, it kind of makes sense: Google said its stock would go up, so people naturally bought, but Google was joking, so they lost money, and Google had to repay them. But: what did they buy, from whom, and why? There wasn't an active market for Google shares, but employees presumably had options. But if the stock price goes up, the options still allow you to buy it at the lower price at which the options are struck, this being the point of options. If they wanted to exercise early, presumably they had to ask someone in HR how they could do so, which would be a good cue for that HR person to say "There's no reason for you to do that, and why would you want to?" which would settle the whole issue before money changed hands. And, anyway, what loss did Google have to make whole, here? The interest on borrowing money from family members for several days and then paying it back? My guess is that the part where Google played a prank about their valuation did happen as described, but the evocative details don't make sense. (Or there are some less evocative details that could have been relegated to a footnote to make the nitpickers happy. On the nitpicking front, I did decide to forgive the author for talking about a rating—the share of objectives achieved in a given period—expressed as a number between 0 and 1, with 1 referred to as "the integer 1" as the high score. Maybe they did ritually cast it from a float to an int as a celebratory gesture. It is, at least, the kind of thing I can imagine someone arguing about.)

One thing the book foreshadows well is Google's future position in AI: they're talking about it constantly in the narrative, and they even talk about "very, very, very lage language models, much larger than anyone has ever built in the history of mankind." (These were the in modern terms tiny models that powered their translation service.)

Google was also forward-thinking about just how much online behavior they could control. In an interview that almost certainly happened long before Google worried about antitrust, Larry Page said that one reason for the famous I'm Feeling Lucky button was that Google wanted to get people out of the habit of using URLs at all, when they could just search for everything they needed. This is the kind of ambition that's endearing when the founders are still technically taking a break from their PhDs in order to start a little company, but that looks ominously grasping in retrospect. And, of course, it's both: the sunny optimism that leads people to start companies in the first place, and to build the most ambitious version of whatever they were working on, sometimes works out well enough that the people in charge are now running important institutions of the sort that get called to testify before lawmakers and accused of ruining other companies' economics. Executing on the plan implied by I'm Feeling Lucky requires more skill than luck, but the technical and product skill it requires is only weakly related to the political skill necessary to preserve that victory.

Open Thread

Leave a Comment

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

Find a Role

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.

Find Talent with Diff Jobs