More gruel
Link Roundup, 8/19/2018

Link Roundup, 8/19/2018

From Bill Gates’ blog: How the software industry (and others) wreck the basic laws of supply and demand.

From CNBC: More unintended consequences of the Trump’s random tariffs – a spike in used car prices.

From Bloomberg: The drama on Tesla’s board over Elon Musk’s increasingly disruptive behavior.

From Rolling Stone: Matt Taibbi does his thing with Virginia gubernatorial candidate Corey Stewart (“A jowly, anus-faced character in a crew cut who looks like a margarine-fattened version of Sean Spicer”).

From The Miami Herald: Mexico inches toward the always-obvious solution to its failed drug war – decriminalization.


  1. Could somebody elaborate on item 2 from the four things Gates lists? The example of Uber and Lyft doesn’t seem to extend to say, Microsoft Office or iOS. Affiliates that take advantage of a platform, be it an App Store or YouTube or an operating system, aren’t competitors but partners in a synergy.
    I’m assuming there are a myriad of laws to stop your devs from going to your competitors with the product they’ve designed for you, and even if not the danger of losing talent you’ve invested in to rival firms is nothing new.

    1. Uber does the heavy lifting of convincing someone to quit his job, buy a car, and start driving for them (e.g. with marketing, support, signup bonuses, referral bonuses, etc.). Then Lyft can come in and say, “here, just download this app and in 5 minutes you can be driving for us as well”. The asset that Uber developed (a rideshare driver) can be piggybacked on by a rival (Lyft). In contrast, if Ford buys a steel press to make its car doors (a traditional “capital investment” in the old economy), GM can’t just sneak in afterwards and use it at night when the factory is closed. It must buy its own at the same price that Ford paid.

      Of course, Uber could have its drivers sign an exclusive non-compete contract, but that would open it up to arguments that its drivers are employees (the traditional form of creating exclusivity over a human asset), which it doesn’t want to do.

      The difference is that MS and Apple’s affiliates aren’t competing with the platform itself. That is, you’re welcome to create apps to sell on iTunes, but Apple doesn’t allow you to create a competitor to iTunes itself. Whereas Lyft is essentially a competing platform to Uber. Interestingly, Google does allow such a thing on Android. Amazon has a competitor to Google’s Play Store, and there are other 3rd party stores as well. But Google perceives the overall advantage to the Android platform as being big enough to allow these 3rd party stores to exist.

      1. Right. But does the situation with ride-sharing services extend to software in general? The key with the Uber situation appears to have more to do with how employment at Uber is akin to contracting than it does with the assets involved being intangible.

      2. I think I had all of two months experience in the software industry when I sat down with a product manager to ask some questions. A few things had left me confused.

        There didn’t appear to be any “variable costs” in this business, only sunk costs. In other words, it appeared that once the software was written, every additional license we sold was essentially free to us. The price was basically 100% profit, minus operating expenses like sales commissions and marketing. Tech support paid for itself via renewals.

        I’ll never forget the look on the guys face as he explained, “We’re selling air.” I had found the only business I would ever want to work in.

        Those were the days. Software increasingly works like Uber and Lyft, as a service. Since each new license is free, software providers are moving up the value chain, delivering applications+compute as a united platform accessed via Internet, but that commoditizes the whole stack. Customers, instead of buying a permanent license in software, can instead just buy access to the application, delivered right to their browser. It’s still an amazing business model, but not the pure-money-machine model it was fifteen years ago. Once customers develop familiarity with one software platform, it’s relatively easy to pivot to another. And that new software as a service model means the whole stack, from the chips to the code, is replaceable the way you’d replace your cable provider.

        This model is on its way to a desktop near you. Schoolkids are mostly walking around with Chrome-book style devices with little or no local compute capability. They access applications via a browser. The only transaction is the purchase of access licenses from a marketplace of providers. It ends up looking a lot like the Uber/Lyft thing. Still a beautiful business, but not quite as amazing as it was once was.

        Funny side note. In the midst of all this, we’re seeing a revival of interest in real, honest-God desktop computers, but now they’re artisanal. Mostly hand assembled from bespoke parts. The move toward phones and stripped down laptops has robbed users of compute capabilities needed for home gaming and media rooms. So the old is new again.

      3. Chris, I may agree with you regarding the “old software business model”, but software was not the first industry to work that way.

        The variable cost that the pharmaceutical industry pays is a tiny fraction of the fixed cost to create a drug. To be honest, if you look at say, the F-35, the unit production cost is still a fraction of the plane compared to the development cost factored into every unit sell price.

        But you are correct about the “new software model”. Once again, the tera-corps (yeah, we are there now) are hugely involved with the “download the computing power you need” model. Yesterday I used Microsoft Online for an Excel spreadsheet and GoogleDocs for a simple document. On a daily basis a developer buddy of mine spins up and down computing power on Amazon’s AWS system. It is truly Pay as you Go.

        To me, it is terrifying that is the new business model. These tera-corps can promise all they want about data security, but there is nothing stopping them from examining every single bit of data, even copying it into their own storage network.

  2. On the contrary Tesla will now be making money each quarter and if Musk does take it private he won’t have to put up with all the financial ‘experts” continually trying to make him fail in order to make money (AKA rip off other investors)

    The Stock market was one of the big inventions that enabled the industrial revolution
    But that was the first part the – the mechanism for funding companies unfortunately the other part the “tail” – a casino for operators to play with money has been wagging the dog for the last 50 years

  3. EJ

    Musk’s ongoing collapse is a fascinating one. Mocking him has been Left Twitter’s favourite sport for more than a year now, but the sheer speed at which he’s gone from tech messiah to flailing emotional wreck, and how visible it’s becoming even to the casual onlooker, is something I find remarkable.

    Will he be the first tech billionaire to see his company collapse entirely, or even (depending on the SEC) see the inside of a jail cell? One wonders. Two years ago it would have been unthinkable. Now…

    1. Musk follows Tesla along the path of folks who can see the future. Getting others to understand is a bit more challenging.


      Corey Stewart brings to mind the time we spent in the Tube in London last week. Most got along quite well, even though we had little in common. The minority were old, white, and pissed. They were not thrilled with the other folks in the train.

      Rail against the others, even though you are them. A native American friend once told me, “we came a long way to get here”. They were here before us.

      My generation will not fix this. We are the problem. My granddaughter might figure it out. Her generation sees a different future. I pray she is right.

    2. Re: Musk , the man is trying to do too much. No human being can physically run two companies like Tesla and SpaceX, that are essentially that are still in their infancy. He is showing the stress, which was inevitable.

      I think Tesla will inevitably be absorbed, or rather its technology, by one of the established auto manufacturers. I don’t know if Musk’s ego would let him, but selling Tesla and focusing on SpaceX would be the best thing for him, and likely humanity.

      I did some quick historical research, and was surprised to find that Tesla is not even close to the first electric car company, nor even the first to introduce mass assembly for electric cars. I guess Tesla is just the first company to go all-in on electric transportation, and at the time of social media, by far the most visible.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.