Anyone familiar with this blog is familiar with my notes from Peter Thiel’s startup course at Stanford. With over 350,000 readers, more than a million page views, and coverage in the New York Times and Forbes (to name a few), the notes have had a good run. But I’m happy to report that they are only the beginning: Peter and I have decided to write a book called Zero to One.
Why a book? For starters, we can make the notes considerably better. Substantively, we are revising, updating, and expanding on the best parts of the class. Everything else will be improved as well. The prose will be stronger and clearer, without losing the atmosphere of openness and experimental thinking that inheres in the notes format. The design and packaging, too, will provide an entirely different readership experience. In short, we can make a book worth owning.
More broadly, though, books remain important, whether digital or in print. Our view is that some degree of sustained attention, not just brief scanning, is essential for real thinking. Stepping back from the parade of distractions and seriously engaging with a text affords opportunity to think, plan, and create. As Peter has said, meaningful progress requires that we think about the future for more than 140 characters or 15 minutes at a time.
Zero to One will be published by Crown Business, a division of Random House, in March of next year. Stay tuned for lots more details—I’ll try and share as much as I can along the way. Meantime, you can sign up for updates or follow along on twitter/fb here. Or pre-order here!
A long time ago in a galaxy far, far away… I was an attorney.
Indeed, I devoted most of the 1990s to the practice of law, clerking for the United States Court of Appeals for the Fifth Circuit and then litigating for the preeminent Wall Street law firm, Sullivan & Cromwell.
As a young lawyer, most of my billable hours were devoted to legal research and writing. I recall slaving away at my computer, endlessly querying LexisNexis and Westlaw and becoming frustrated with the limitations of crude keyword search and arcane Boolean operators. Indeed, my hack was to spend many days and nights in the library reading cases in printed books to track down the key facts and subtle distinctions that the “computer” could not grasp.
Of course, most of the world of technology has advanced since those dark days. But not legal research. Until now.
Fixing legal research is a major task. To start, it requires a scalable method of extracting meaning from millions of cases, not just adding a more advanced search engine on top of the text.
Judicata is developing an intuitive search technology that groks all of the facets of legal precedent. In a matter of moments, their software helps a lawyer retrieve everything she needs, comprehensively, accurately, and painlessly. By early next year, the team will ship the new tool of choice for California’s 180,000 lawyers.
According to Crunchbase, legal technology attracts fewer investment dollars than any other sector, and perhaps for good reason; the problem is difficult, and most companies are taking incremental approaches. Yet there is a lucrative market awaiting the right team with the right approach: the two industry giants generate over $2 billion annually from their legal research products, and the largest 100 law firms alone generated $70 billion of revenue last year.
Real innovation is possible in legal technology, and it is on the horizon. We at Khosla Ventures are excited to be working with the Judicata team to prove it.
In 1811, when Cornelius Vanderbilt was 17, he borrowed $100 from his mom to buy a small sailboat. He figured he could make some money by ferrying goods and people around New York Harbor. He was right.
When the War of 1812 broke out, Vanderbilt’s competition nearly vanished (presumably, few American transporters were keen on operating in British-infested waters). The demand for effective transport, though—particularly military transport—increased dramatically. Vanderbilt, who quickly acquired the nickname “Commodore” for his prowess on the water, was all too happy to service the need and profit handsomely therefrom.
Vanderbilt, of course, was the sort of guy who thought seriously about the future, and the future, he thought, was steam power. So in 1818 he sold his fleet, leased a steamship called Bellona from a guy named Thomas Gibbons, and began to operate his ferry business 2.0.
But there was trouble on the water. The New York legislature had seen fit to grant a monopoly on steamboat service to a couple of guys named Fulton and Livingston. Some operators, like Gibbons, respected the edict and stayed out of the water. Others, like Aaron Ogden, cowed and paid the Fulton-Livingston partnership for an operating license. But Vanderbilt was made of different stuff. He just wanted to build a great business. What good are rules when they stand in the way of building great businesses?
Unsurprisingly, suits were filed. (Ogden was the plaintiff in the one you’ve probably heard of.) Interestingly, though, this didn’t seem to matter very much. Initially, Vanderbilt paid the litigation no mind; he continued to provide excellent service and ruthlessly undercut his competition on price. Equal parts sword and shield—he employed a “crew of shoulder-hitters, ready for battle” to ensure orderly moorings at competitor’s docks,1 while also deflecting criticism and developing a Robin Hood-ish mythology—Vanderbilt insisted on forging his own future. You might be aware that Jay-Z just executive produced Baz Luhrmann’s Gatsby; so long as we’re anachronistically weaving Hova lyrics into montages of the long-dead nouveau riche, take a moment to imagine Vanderbilt, as his marine hoplites take control of a pier, blasting:
And government, fuck government, niggas politic themselves.2
The end of the legal battle came in 1824, when the Supreme Court heard the case and ruled for Vanderbilt’s side. (Vanderbilt, as merciless in court as he was in business, had helped his cause by hiring Daniel Webster—think Ted Olson and David Boies rolled into one—to represent Gibbons.) Doctrinally, the case was quite important, but that is the stuff of AP US History and 1L year of law school. What matters here is that Vanderbilt was venerated:
“We owe to him,” said a prominent citizen, “the freedom of the seas as applied to us locally.”3
I think this story is pretty cool in its own right. It’s even cooler, though, to the extent it can help us understand the present. Does the Vanderbilt steamship ordeal remind you of anything more… familiar? Say, much of Silicon Valley right now? I’ll let someone else write the manifesto about how technology is and will likely continue to outpace physical-world regulators and solve problems the government can’t. But cf. Uber/Airbnb/Taskrabbit/Exec/Crowdflower/Turk/3D Printing. It’s hard not to notice that CS can be a powerful mechanism to route around inefficiency and unlock a lot of value.
Of course, disruption is risky. People don’t like to be disrupted. Aaron Ogden certainly didn’t. Neither, apparently, do the bureaucrats in DC who are coming after Defense Distributed, ostensibly because they feel weak and techno-libertarianism scored too many points over the weekend, or something. The best path is usually one that avoids head-on confrontation. But still—very often, it’s messy and complicated where the rubber hits the road. So what should we do then this happens? Play by all the rules? Ask for permission? Or just build something great? To ask the question, hopefully, is to answer it. WWVanderbiltD?
Over the last year or so, I’ve had the pleasure of watching my good friends Kyle and Dan build Leap, which, as they bill it, is “a better bus service for San Francisco.” The idea is simple: the city’s MUNI bus system ($2/ride) is slow, overcrowded, and leaves much to be desired.4 But biking (free) isn’t for everyone, and cabs ($20) are expensive. What if we could relieve the MUNI’s load by bringing the private shuttle service that Google and Twitter employees enjoy to… everyone? What if anyone with a smartphone could instantly buy a pass and streamline their commute on a bus with wi-fi, air conditioning, and a comfortable seat? Well, please meet Leap ($6), which launched this week with a line from the Marina to Downtown SF.
It’s always fun to watch your friends start new ventures. It’s also fun to see really good products get built. Throw in the delightful parallels to the Bellona line and it’s not hard to imagine Kyle and Dan and company as a couple of proto-Vanderbilts, just trying to get people from point A to point B in a better way.
May the streets of San Francisco be their New York Harbor.
Stewart H. Holbrook, The Age of the Moguls, 13 (1953). ↩
Jay-Z, Decoded, 214 (2011). Note the esoteric use of “politic,” glossed in p. 215 n20: “I wrote this at a time when I felt the government was irrelevant to the ways we organized, resolved conflict, and took care of ourselves. “Politic” is slang for the kind of talk that works things out.” ↩
Holbrook, supra, at 13. ↩
Others have written—some quite critically—about the government’s decision to charge Dzhokhar Tsarnaev with “unlawfully using and conspiring to use a weapon of mass destruction.” Colloquially, most people have come to associate WMDs with nuclear, biological, or chemical weapons.1 But 18 USC § 2332a(c)(2)(A) says that “any destructive device as defined in 18 USC § 921" is a "weapon of mass destruction." Section 921(a)(4)(A) in turn, says that bombs, grenades, and rockets, among other things, are “destructive devices.”2
Since the IEDs used in the Boston attacks are clearly “bombs,” the § 2332(a) charge is pretty straightforward. People can argue about whether that statute is too broad, and whether other laws against, say, murder, would suffice. But what caught my eye in § 921(a) was the grenade and rocket stuff.
It turns out that another way to say “grenade launcher” or “rocket launcher” is “flare gun.” The government knows this and doesn’t really care about your flares, so there is an exception in 18 USC § 921(a)(4) (and in 26 USC 845(f)(3), an identical provision in the NFA) that says that a “signaling, pyrotechnic, line throwing, safety, or similar device" is not a destructive device. (Presumably, the ATF would say that a signaling device that is used to launch grenades is no longer a signaling device.)
Whether certain kinds of launchers are WMDs thus depends on what sort of ammunition you use or plan to use.3 Firing unregistered grenades or rockets obviously triggers § 921(a)(4)(A) and an avalanche of criminal liability. That makes sense.
So what’s between a grenade and a flare? Well, according to the ATF, “cartridges containing wood pellets, rubber pellets or balls, or bean bags" are "'anti-personnel' ammunition,” so 37/38mm launchers containing such loads are d̶e̶s̶t̶r̶u̶c̶t̶i̶v̶e̶ ̶d̶e̶v̶i̶c̶e̶s̶ WMDs. That’s right: launcher + bean bag round = WMD.
DYODD and check your state laws, but it appears that in some jurisdictions one can order a pretty serious launcher and have it shipped right to his doorstep; absent “anti-personnel” ammo, the federal government doesn’t even consider it a gun. Just stick to fireworks and take care not to possess (let alone fire) any bean bag rounds if you don’t want to spend “a term of years, up to life” in Leavenworth.
By this definition, of course, there were tons of WMDs in Iraq. Not that definitions must be consistent across domestic/international or colloquial/technical contexts, but that’s at least worth thinking about. ↩
Under federal law, 40mm launchers are always destructive devices, and thus must be registered to be lawfully possessed. Launchers < 40mm, as far as I can tell, are different. ↩
Well, I am officially a lawyer! Yesterday the Judicata team went over to the James R. Browning Courthouse, where Alex Kozinski, the Chief Judge of the U.S. Court of Appeals for the Ninth Circuit, swore me in to the California Bar. (Or did he? Technically, I affirmed an affirmation instead of swearing an oath. More on that in a bit.)
First, we got a private tour from Kathleen Butterfield, one of the Court’s staff attorneys. The Courthouse is, in a word, incredible. I think that most of what we saw is open to the public during the Court’s bimonthly public tours; if you’re near San Francisco, please, take my advice and attend one. (You might ask if or when Kathleen is leading a tour—she is terrific.)
(1) So much Italian marble it makes the Hearst Castle look budget:
(2) The bar—literally. Before law schools existed, would-be lawyers would study the law under another lawyer’s supervision. Getting admitted to the Bar involved standing behind the bar with your sponsor and fielding a bunch of questions from the judges. Get enough right and you’d be permitted to—wait for it—literally pass the bar.
(3) The bullet hole from the Hindu-German Conspiracy Trial. In 1918, not five feet from where we held our ceremony, a defendant shot and killed his co-defendant and was then promptly shot to death by a U.S. Marshal. (Amazingly, no mistrial occurred; everybody was found guilty the next week.) You can still see the damage caused by one of the bullets when it hit the judges’ bench—check out the aberration in the tilework, just to the right of the seam in the marble:
After the tour, we hung out in Courtroom One until Judge Kozinski freed up.
After the Judge came in and met the rest of the team, I asked if he’d mind if I chose to affirm rather than to swear. Legally, there’s no difference. Swearing is traditionally perceived to have a religious component to it, whereas affirming is completely secular. This is a pretty mainstream option—the U.S. Constitution explicitly follows every “Oath” with “or Affirmation,” and the official California Bar incantation reads “swear (or affirm)”—but I’d bet that it’s seldom exercised. (Of all my lawyer friends, I know just one who affirmed, and we had discussed it beforehand.)
Why would anyone be so fussy? Naturally, atheists or radically liberal First Amendment zealots tend to be quite interested in keeping things as secular as possible. But even theists have their reasons:
But I say unto you, swear not at all: neither by Heaven, for it is God’s throne;1
But let your communication be ‘yea, yea’ or ‘nay, nay’; for whatsoever is more than these cometh of evil.2
Personally, I chose to affirm because (a) I could, and (b) it seems cooler. Presumably, some of our forefathers argued long and hard to win for us the right to affirm. Why not throw them a cosmic wink? Plus, if it was good enough for Franklin Pierce, it’s good enough for me.
Of course, the Judge was cool with it, and we got it done:
Afterwards, we sat down at the Appellant’s table to chat about Judicata and legal technology. For those of you who don’t know, Judge Kozinski is a pretty tech-savvy guy. After we discussed Judicata’s version of man-machine symbiosis, he dialed back the clock and dazzled us with stories about when he used to program in Fortran on IBM punch cards.
The night ended with dinner at a nearby restaurant. Naturally, the Judge and the whip-smart Ninth Circuit clerks that joined us were delightful company.
I’d like to thank Judge Kozinski and everybody at the Court who made our visit especially memorable yesterday!
I did a keynote interview last night at kbs+ ventures for their new book's launch event. Since I talked a bit about CS183 and the Thiel Fellowship program, a few people came up to me afterwards and asked different versions of the same question: if I wanted to be an entrepreneur, why did I get several degrees from Stanford instead of dropping out?
My quick answer was that it’s important to avoid blanket statements about education and entrepreneurship. Certainly many successful entrepreneurs have name brand college degrees. But many don’t.
Last I looked, Wikipedia’s list of college dropout billionaires is 31 people and counting (and only one of them is a drug lord). I thought that was a shockingly high number, as most of us only know the most famous three or four. If we wanted to talk about dropout millionaires, there are so many that we’d probably need scientific notation.
One of the reasons for this is that the market doesn’t necessarily wait 4 years for you to get your BS or 6 or 7 years for your PhD. In 2003 and 2004, Mark Zuckerberg had a huge advantage in that he was working furiously toward something he sensed was important while his peers were still locked into school. Starting Facebook in 2007 would not have worked.
One key distinction is between businesses that require a lot of specialized domain knowledge and businesses that don’t. Often, this tracks the distinction between enterprise/B2B and consumer models. Bright, well-adjusted 18- or 19-year-olds can develop the kind of social insight that’s at the core of many great products, maybe even better than older folks can. If their engineering skills are adequate, they can build the vision and more or less take over the world. This isn’t to say this is easy, of course—only that it’s very possible.
Facebook was one example. Another may be Gumroad, a novel e-commerce company run by 20-year-old Sahil Lavingia. If you aren’t already familiar with it, Gumroad enables anyone to sell something online in a matter of seconds. If you want to sell an e-book, for example, you create a product listing and get a unique link that you can share throughout the web. If, as Sahil says, Gumroad “becomes a thing,” he will have succeeded in turning all of Facebook and Twitter into a global online marketplace, i.e. in building a billion dollar company.
Other businesses require a great deal of domain-specific knowledge, which often entails specialized education. For example, my company, Judicata, builds radically better legal search and analytics software for lawyers. This requires great engineers, but it also requires great lawyers who deeply understand how the law works. For better or worse—and I actually suspect for worse—one almost invariably needs to go to college and then law school to become a lawyer. At the very least, being in law school affords one a structured opportunity to learn how to think about the law.
For some ventures, getting a technical or professional education is unquestionably the right move. For others, college is absolutely the wrong move. More interesting than the drop out vs. not question, I think, is the set of questions that aims at unpacking what a college education really is. To really understand the nexus between education and entrepreneurship, we’d be better off starting there.
In most respects, U.S. Attorney Carmen Ortiz’s statement about the Aaron Swartz prosecution is unremarkable. It’s more or less the standard fare that one expects from government officials who unexpectedly find themselves on their heels.
The most striking thing about the statement is the writing itself, which is terrible. Ortiz begins:
As a parent and a sister, I can only imagine the pain felt by the family and friends of Aaron Swartz, and I want to extend my heartfelt sympathy to everyone who knew and loved this young man.
What is the first clause doing here? Presumably it is either intended to establish empathy or sympathy. Empathy seems odd because Ortiz extends “sympathy” two sentences later. But sympathy doesn’t require mentioning Ortiz’s own family, unless she means to syllogize: “I would be upset if one of my family members killed himself. Aaron killed himself. Therefore I imagine that Aaron’s family is upset.” Much better is to avoid the token contrivance and say something like:
I want to extend my heartfelt sympathy to everyone who knew and loved Aaron Swartz. All of us at the U.S. Attorney’s Office deeply regret that Aaron chose to take his own life.
I know that there is little I can say to abate the anger felt by those who believe that this office’s prosecution of Mr. Swartz was unwarranted and somehow led to the tragic result of him taking his own life.
This isn’t that bad. It sort of feels like it’s missing a sentence at the end. The bigger problem is that “somehow led” is too snarky. Ortiz is basically saying that anyone who believes that Aaron’s suicide was related to his prosecution is an idiot. The problem is that most people do believe that these events are related, which seems quite reasonable absent evidence to the contrary. Ortiz is pushing back on causation where she should be pushing back elsewhere instead. The weak line is “We didn’t do anything that caused Aaron to kill himself.” The strong line is “We did everything right. It sucks that Aaron killed himself. Not our fault.”
Ortiz’s second paragraph begins as haphazardly as the first one ended:
I must, however, make clear that this office’s conduct was appropriate in bringing and handling this case.
This sounds weak and defensive in the worst way. If you didn’t do anything wrong, why not take ownership and avoid the passive voice? She continues to wax Rumsfeldian:
The career prosecutors handling this matter took on the difficult task of enforcing a law they had taken an oath to uphold, and did so reasonably.
At one level, she is defending her prosecutors. On another level, it seems like she is throwing them under the bus. Also, no one wants to hear how hard your job is when you mess up (or if you are perceived to have messed up).
Ortiz then spends the rest of the paragraph aimlessly listing off some facts that she hopes readers will find mitigating. Why she doesn’t use an unordered list is beyond me. (The federal government, after all, knows how to use bullet points.) Before we get there, though, consider whether something like the following would make a better second paragraph:
I have been involved with this case since its inception. I have reviewed the record at length over the past few days. My conclusion is clear: I am absolutely certain that our conduct was appropriate in bringing and handling this case. Prosecutors make odd scapegoats in situations such as these. Our job is to enforce valid laws fairly and effectively. In this case, that meant bringing charges against Aaron. We do not pass judgment on the wisdom of laws that Congress chooses to enact. And, unfortunately, we cannot prevent the small percentage of criminal defendants who elect to commit suicide from doing so.
Ortiz’s fact section is her worst writing of all. Consider the following excerpt:
That is why in the discussions with his counsel about a resolution of the case this office sought an appropriate sentence that matched the alleged conduct – a sentence that we would recommend to the judge of six months in a low security setting. While at the same time, his defense counsel would have been free to recommend a sentence of probation. Ultimately, any sentence imposed would have been up to the judge.
That is atrocious. I don’t think it needs any more explanation than that. If you disagree, you are wrong. If you think that this horror show is necessary legalese, you are wrong. This is not a real estate contract. It does not need to read this way. Whether it’s bad enough to do any damage is unclear, but it is certainly not persuasive writing. AUSAs are typically pretty good lawyers who write well. I seriously question whether Ortiz had any of her people read her statement before releasing it. (Another possibility, as @DanGoldin suggests, is that too many people worked on it. Either Ortiz has zero proofreaders or an army of them.)
In any case, here is how one might retool the facts section:
- We recognized that there was no evidence that indicated that Aaron acted for personal financial gain.
- We recognized that Aaron’s crimes did not warrant the maximum punishments authorized by Congress.
- We never sought—or told Aaron or his attorney that we intended to seek—the maximum penalties in this case.
- When we discussed settlement options with Aaron and his lawyer, we told them that we would recommend a sentence of six months in a low security facility if Aaron agreed to plead guilty. Aaron declined.
- Judges, not prosecutors, decide sentences. All prosecutors can do is recommend a particular sentence. Note that even under a plea bargain, Aaron’s counsel would have been free to recommend and advocate for an even lesser sentence than six months.
The statement ends with:
As federal prosecutors, our mission includes protecting the use of computers and the Internet by enforcing the law as fairly and responsibly as possible. We strive to do our best to fulfill this mission every day.
That’s not so bad. But could she have ended stronger?
While anger is understandable in the wake of tragedy, anger at prosecutors, at least in this case, is misplaced. If people believe that the law should be changed, certainly they should contact their representatives in Congress. In the meantime, as federal prosecutors, our mission is to enforce the law as fairly and responsibly as possible. That is what we did in this case, and we will continue to do just that.
I don’t mean to suggest that my edits make the statement appropriate. I did them in 20 minutes on the Caltrain this morning. What I do mean to suggest is that there is lots of room for improvement. Since this was surely a statement worth getting right, I can’t help but wonder what went wrong.