Questions · Patrick Collison


Some questions that I find interesting. I've omitted some that are related to Stripe. (Pointers to interesting readings on these topics are always welcome.)

  • Why are certain things getting so much more expensive?

    Spending on healthcare in the US is up 9X in real terms since 1960. K12 education spending in the US has increased by 2-3X per student per year since 1960. The cost of college in the US has more than doubled (again, in real terms) since 1984. Growth in everything from construction costs and childcare costs is significantly outpacing inflation. Lots more at SSC and from Tyler.

    What's going on? Why are we seeing dramatic declines in costs in many areas of the economy and such steep increases in others? How much of the cost growth is unmeasured improvement in quality and how much is growing inefficiency? How should one predict a priori whether a sector will exhibit increasing or decreasing costs relative to inflation? What do we do about it all?

  • Why do there seem to be more examples of rapidly-completed major projects in the past than the present?

    The Empire State Building was built in 410 days. The Lockheed P-80, the first jet aircraft deployed in the Air Force, took 143 days from project initiation to first deployment. The Apollo Program lasted 9 years from initiation to moon landing. While a lot happened in the US during World War II, it's easy to forget how short the period in question was: American involvement lasted 3 years 8 months and 23 days.

    Meanwhile, a BART extension is delayed more than a year because the wrong networking equipment was installed. (The 16-mile extension will cost circa $2.3 billion and have taken around 7 years.)

    Is the sense that there are fewer contemporary examples of rapid progress justified? If so, what's going on?

  • Why is US GDP growth so weirdly constant?

    If you look at log US GDP over the past 150-or-so years, it is very weirdly smooth. Why? What determines the slope? Would it be correct to conclude that "almost nothing will affect the economy over the long run"? This phenomenon may even extend back further in the US. But it's not like nothing matters; GDP growth between countries does vary a lot, in both the short and the long run. So... what gives?

  • How do you ensure an adequate replacement rate in systems that have no natural way to die?

    Schumpeter claimed that the problem of capitalism is not how capital is allocated to existing structures but how structures are created and destroyed. Systems and institutions inevitably get stale or become less effective but also work very hard to survive. (See also: the institutional imperative.) Bankruptcy takes care of senescent businesses. But how do we get a sufficient replacement rate in systems and institutions that aren't naturally subject to extinction processes?

  • How do we help more experimental cities get started?

    It seems that the returns to entrepreneurialism in cities remain high: Hong Kong, Singapore, Dubai, and others, have improved the lives of millions of people and appear much more contingent than inevitable. Maybe there could be far more of them. Beyond the direct benefits, city-sized areas enable regulatory experimentation that may, in turn, affect much larger regions. The Shenzhen special economic zone was the first step in China's broader economic liberalization. (In addition to the books about the aforementioned cities, some related reading.)

    How can we encourage the creation of more cities and more experimentation in their rules?

  • How do people decide to make major life changes?

    Most days, people don't decide to change their lives in big ways. On a few days, they do. What's special about those days? How much is it about the stimulus versus their own inner state?

  • Why are there so many successful startups in Stockholm?

    London and Paris have surprisingly few successful tech startups for their size. Stockholm, a city of less than 1 million people, has Spotify, King, Klarna, iZettle, and Mojang, all valued at more than $1 billion. What's true of Stockholm that isn't true of other European cities? (Provo may be the US Stockholm and some similar questions apply there.)

  • Is Bloom's "Two Sigma" phenomenon real? If so, what do we do about it?

    Educational psychologist Benjamin Bloom found that one-on-one tutoring using mastery learning led to a two sigma improvement in student performance. The results were replicated. He asks in his paper that identified the "2 Sigma Problem": how do we achieve these results in conditions more practical (i.e., more scalable) than one-to-one tutoring?

  • How can we better understand the dynamics of progress in science?

    We invest an enormous amount of time and money in science. Cumulative progress in science will probably do more than anything else to determine the shape of our collective future. In order to ensure that we're spending enough—and doing so on the right things and in the right ways (maybe we should have more independent institutes, or more funding of promising early-career scientists, or...)—how should we think about measuring the results?

  • Will end-user applications ever be truly programmable? If so, how?

    Emacs, Smalltalk, Genera, and VBA embody a vision of malleable end-user computing: if the application doesn't do what you want, it's easy to tweak or augment it to suit your purposes. Today, however, end-user software increasingly operates behind bulletproof glass. This is especially true in the growth areas: mobile and web apps. Furthermore, not only is it getting harder to manipulate the application logic itself, but it's also becoming harder to directly manipulate your data. With Visual Basic, you can readily write a quick script to calculate some calendar analytics with Outlook. To do the same with Google Calendar is a very laborious chore.

    End-user computing is becoming less a bicycle and more a monorail for the mind.

    As a consequence, we need ever more domain-specific software. Rather than use universal tools for handling charts and for manipulating data, we tend to use separate analytics packages for every conceivable application. This is not all bad. Domain-specific tools can maximize ease-of-use and help amortize the cost of complex, specialized functionality. Sublime's built-in ⌘-T works better than every third-party Emacs package. Still, despite these benefits, the popularity of macros and browser plugins strongly suggest that users are smart and want more control.

    Should we just give up on our earlier visions of empowered users or is a better equilibrium possible?

  • What's the successor to the book? And how could books be improved?

    Books are great (unless you're Socrates). We now have magic ink. As an artifact for effecting the transmission of knowledge (rather than a source of entertainment), how can the book be improved? How can we help authors understand how well their work is doing in practice? (Which parts are readers confused by or stumbling over or skipping?) How can we follow shared annotations by the people we admire? Being limited in our years on the earth, how can we incentivize brevity? Is there any way to facilitate user-suggested improvements?

  • What's the successor to the scientific paper and the scientific journal?

    Are LaTeX'd papers and paid journals really the best way can do? Peer review and modern scientific publishing are quite recent phenomena. In different ways, Distill, the arXiv, Fermat's Library, and Sci-Hub hint that improvement might be possible.

  • What's the right way to understand and model personality?

    Since the 1980s, the five-factor personality model has gained a great deal of traction. It is ostensibly applicable to a great deal of life. What are its limitations and in which situations should we use it?

  • Could there be more good blogs?

    It seems that they heyday of of blogging is passing. If so, that's unfortunate. Blogs can be a remarkably efficient mechanism for disseminating ideas and facilitating discussion and debate. (E.g., Sumner's arguments for NGDP targeting) Twitter is good, too, but there's lots that blogs are great for that Twitter can't replace.

    Part of the problem with blogs is that they're less rewarding than Facebook and Twitter: your post may perhaps get some thoughtful responses but it doesn't get immediate likes. And part of the problem is, of course, that writing a good post is much harder than writing a witty tweet.

    Are there incentive structure tweaks that yield more good blogging?

  • Why are programming environments still so primitive?

    In different ways, Mathematica, Genera, and Smalltalk put almost every other programming environment to shame. Atom, Sublime Edit, and Visual Studio Code are neat, but they do not represent a great improvement over TextMate circa 2007. Emacs and Vim have advanced by even less.

    Why can't I connect my editor to a running program and hover over values to see what they last were? Why isn't time-traveling debugging widely deployed? Why can't I debug a function without restarting my program? Why in the name of the good lord are REPLs still textual? Why can't I copy a URL to my editor to enable real-time collaboration with someone else? Why isn't my editor integrated with the terminal? Why doesn't autocomplete help me based on the adjacent problems others have solved?