The Boston Globe on breaking up Big Tech falls into the trap of tech exceptionalism

Originally published at:


Well, and simply because it’s new. The arrow of capitalism points toward monopoly, and when new commercial activities appear, they inevitably grow that way until someone notices that a harmful monopoly has arisen (either as a single large company, or as a guild of some kind).

If I invented a machine that made pooping unnecessary, no one would be calling for my business to be broken up on day one. But decades later, when no one owned toilets any more, my monopoly over the means of elimination would be an obvious problem.

Short of full communism, that’s just how things work. The test of a good regulatory regime is not that it prevents this in the first place (which really would stifle innovation). The test is whether you can intervene at just the right time. Part of which is recognising that yesterday’s innovator is tomorrow’s monopolist, and not feigning surprise every time it happens.


As a techie, I can honestly say a breakup of any given large tech company would just partially reverse the progress of their constant aqcuisitions. It would be fine.


Precisely. There’s nothing special about machine learning. It’s nothing more than a fancy kind of statistical analysis. I prefer to call it “curve fitting” instead, to remind my research team (who are busy using it for all sorts of interesting things) to remain humble about what they do. Lots of where places it’s applied in the software industry, a linear regression would work just as well.


I don’t actually mind when machine learning is called machine learning rather than just statistics. That’s fine. And even a lot of textbooks on machine learning start out with regression, explaining that that can be seen as the simplest form of ML. What I hate is when it is called “Artificial Intelligence”. That’s just hyperbole.


I don’t think this is the case. There is actually a problem with companies first to market. They could have killer ideas but wrong on the details (like MySpace) or be crushed by later market standardization (like McLean and Matson with shipping containers) or didn’t quite have the technology yet (like many of the earliest car or airplane manufacturers). Monopolization isn’t automatic - it’s worked for as a goal in an of itself.


Or maybe they look the landscape now, and think well that was a waste of time and money.

Well I guess you are to be complimented for giving one whole parenthetical word to hinting at the existence of a solution that doesn’t require massive government intervention.

1 Like

That’s why I said “new activities” and not “new companies”. But whether the monopoly is the original innovator (AT&T), or someone else who ate their lunch early on (Microsoft) or a cartel of nominal competitors (pre-industrial guilds), it’s nearly the same result as far as the rest of the world is concerned.


One of Cory’s better pieces in a bit, IMHO.


Well what would a Harvard professor know about economics? We should ask the guy that runs Chik-Fil-A, or maybe some Kardashians.


Lets do it in time order. The oldest most purely evil monopolies first. Something like this (I’m sure I’m forgetting many):

  1. Financial / Banks
  2. Fossil Fuels / Energy
  3. Manufacturing
  4. Agriculture
  5. Telecos
  6. Pharmaceuticals
  7. CableCos
  8. Facebook
  9. Google

Religions have for a long time been the most ruthless monopolies, and they often work a lot like profit maximizing businesses too.




Well there are two things that need to happen (IMO)

  • Admit that economic harm isn’t the only harm that can happen to the public
  • based on the first - admit that there exists a binary choice if a company is an monopoly that has brought harm - accept heavy regulation or to be broken up.

A good start would be to realize higher prices isn’t the only form of harm. Come to think of it, what harm isn’t economic? Even regulatory capture has a price tag.


I mean, they’re not that new. Facebook: 15 years old. Google: 20 years old. Amazon: 25 years old. (And yeah, sure they didn’t all get to be problematically large instantly, but, for a historical example, the move to break up the American Tobacco Company was taken 17 years after founding.) It shouldn’t take decades to realize there’s a monopoly situation, normally.

Granted, to some degree, we’re still figuring out what the negative impacts are of having monopolies in these industries, but it should have been long clear that they were monopolies.

There is, on the whole, an increasing move towards, and tolerance of, monopolies, even with “old” industries and businesses that they’re benefiting from:

Surely not. I haven’t looked them up lately, but I’m sure Infoseek, Yahoo, Lycos, Looksmart, Excite, AltaVista and Ask Jeeves are still industry titans!


What harm isn’t economic is a point that can be argued around to keep nothing changing - While it might make better law, letting pedantic arguments halt all progress is a major part of what is currently wrong with this country. Just look at the ‘bump stock’ ban and the endless arguments about how bump stocks aren’t technically breaking any rules. They are all correct btw - bump stocks under existing law should be legal - but they also should be banned - and here we are - where ignoring the pedantry is the only way to move an issue forward.

1 Like

So…how about we start by tackling the oligopsony (a very fun word to say) in high tech employment by banning (or sans government unionizing and negotiating) the noncompete contract that prevent these experts from working in their field for years after departure?

You will never get a second capable search engine if all of the search engineers have to continue working for Google or learn a new expertise

Also…why aren’t any of these companies pursuing national champion status to prevent being broken up? That shit isn’t given away just because you got big, it needs to be earned. Amazon could dip it’s toes into geopolitics by having a search tab for goods produced by unions, for instance (If it proved popular, it would benefit the US and every one of our allies, but absolutely exclude China since they, hilariously, don’t have unions.

1 Like

This article and attending conversation are all very fascinating and I plan to spend more time exploring the various points made when I have the time to spare but for now I just want to hear more about this anti-poop machine.


I didn’t in any way mean to suggest the current regulatory regime is doing a good job. Amazon, for example, should have been setting off alarms for the last half-dozen US election cycles. For its first decade, it’s understandable that people would find it convenient and refreshing, but it’s not all that different from Wal-Mart, except in degree. By the late aughts it was obvious how and why Amazon was going to become a big problem.

Google and Facebook really are more complicated, though. Their business is selling ads, and they do have a duopoly in that business. But it’s their cultural consequences that people care about, and it’s not clear that “antitrust” is the right frame for thinking about those concerns. I strongly suspect that having the state take charge of Facebook, in particular, would reify Facebook’s undeserved, toxic role in civic life – entrenching the problem – when there is a better and easier option to just let it die.

(Or perhaps I should say encourage it to die, with robust, general privacy laws)