The Winner-Take-All Trend In Tech (newyorker.com)

An anonymous reader writes: A pair of articles about the tech industry serve to highlight a growing trend. First, Om Malik writes in The New Yorker about the failure of ride-sharing company Sidecar, backed by Richard Branson, and how it's one more example of the winner-take-all tendency with modern tech firms. "This loop of algorithms, infrastructure, and data is potent. Add what are called network effects to the mix, and you start to see virtual monopolies emerge almost overnight. ... The more we use it, the more data we give the company, and the more it is able to control where we turn our attention."

The second article is from Jacques Mattheij, who notes a different side of the trend toward one winner and a whole lot of losers: unnecessary reliance on cloud-based components to force vendor lock-in. "In many of these cases if you look a bit more closely at what is being sold you'll realize that these are just instances of a business-model that was grafted on as an afterthought onto something that would have worked really well stand-alone but where the creators weren't happy with a one-time fee from potential buyers." Companies who hit it big early can't help but stay dominant if they force users to rely on their servers.

Comments

That is why standards are so useful

[Lennie]

That is why standards are so useful and we need to keep making them.

When winner-takes-all happens, it's better to have the standard be the winner. Not some company.

Because when winner-takes-all happens with companies you get a monopoly and abuse of that monopoly (possibly from the pressure of the stock market demanding better and better numbers).

Re:That is why standards are so useful

[Lennie]

But. But. But.

The Free Market doesn't WORK THAT WAY!

Does it? Does it?

A lot more negative way to look at it is this: companies that fear to not be the winner will help build standards to prevent other companies to be the winner.

Look at this:
https://www.opencontainers.org...

A company like VMware might be afraid Docker could be the winner because of the network effect of Docker, a company like Microsoft might be afraid Google could be the winner. CoreOS might be afraid Docker could kill their business. Who knows. Maybe I'm wrong.

But whatever the motive of the companies or the individuals, they are making standards and even open source code: https://runc.io/

I personally prefer winner-takes-all standards over that over winner-takes-all companies.

The market can deal with the rest of the problem: making products and services around standards.

Basic economics

[Britz]

I know that economics has a bad reputation. Rightly so. Keynes doesn't make sense, but seems to work and the crazy libertarians make a lot of sense in theory, but got us 2008.

Yet some models and explanations are solid and work. One of them is the concept of the natural monopoly. I would argue that Facebook, Microsoft (Windows and Office) operate in markets with natural monopolies.

Markets with a natural monopoly don't work well in a market economy. They either need to be heavily regulated or simply taken over by the state and out of the private industry. Both models aren't ideal.

But Microsoft isn't regulated. Facebook neither. This happens because of globalization. Nation states would need to regulate them, but since those companies operate on a global scale, there is little interest by national regulators to step in. Since there is no international regulator, there is no regulation. Hence these companies are free to exploit their natural monopoly.

Re:Basic economics

[Kjella]

Yet some models and explanations are solid and work. One of them is the concept of the natural monopoly. I would argue that Facebook, Microsoft (Windows and Office) operate in markets with natural monopolies.

Yes, but for different reasons. Software is in general a natural monopoly. There's a mostly fixed development cost, hardly any marginal cost. That means there is a strong cost incentive to reduce the number of implementations until there is only one, which is the definition of a natural monopoly. On the bright side there are two alternatives, a proprietary solution monopolized by one company or an open source solution shared by everyone. Both are theoretically efficient since there's no reinventing of the wheel. Of course once you introduce a few more real world conditions like whether there's is really one product to fit every need, how the feedback cycle from profit/benefit to new development works etc. it's a bit more complicated.

Social networks on the other hand have a huge network effect. That is to say, the value of an empty Facebook is almost none. Even they gave away all the source, said you can build your own site but not take any accounts or anything posted on Facebook, it would be almost worthless. So while software is a natural monopoly on the cost side, Facebook is a natural monopoly on the benefit side. Like the phone network, if you couldn't call someone with a different provider it doesn't matter even if the technology is trivial. That one is much harder to solve because you can always build better software, open source carries on even with almost no market share. But even if you have a better service, getting all the users to jump sides is incredibly hard. It's like how I'm sure it'd be easier if we all agreed to drive on the left or the right or 50Hz vs 60Hz or 110V vs 220V but the costs involved are absolutely massive compared to the gains.

Re:Basic economics

[NostalgiaForInfinity]

I know that economics has a bad reputation. Rightly so. Keynes doesn't make sense, but seems to work

Keynesianism does make sense, unfortunately it doesn't work in practice. For example, empirical evidence is pretty strong that the Keynesian multiplier (how many dollars the economy gains from each dollars spent by the government) is less than one. Progressives and Democrats on the one hand complain about crony capitalism and big corporations, and on the other hand hand out trillions of dollars to such corporations in stimulus spending and bailouts. And when these programs don't yield the promised results, they say that they should have spent even more. How stupid do you have to be to believe this crap?

and the crazy libertarians make a lot of sense in theory, but got us 2008

How can "libertarians" be responsible for anything in our economy? We haven't had anything even remotely resembling libertarian government for more than a century and government regulations have been steadily increasing. The few areas where we have had "deregulation" and "privatization" (e.g., telecoms, airlines) have not resulted in anything like a free market (although they have still been beneficial). Most deregulation and privatization by Democrats and Republicans have themselves been tied up with corporate interests and crony capitalism, something both parties are deeply beholden to.

Yet some models and explanations are solid and work. One of them is the concept of the natural monopoly.

There is little evidence that natural monopolies exist in any economically meaningful sense. That is, if you define your market sufficiently narrowly, you can claim that some company has a "natural monopoly", but there is no reason to believe that your definition of "market" is economically relevant. For example, if you define the "desktop PC" as a market, Microsoft has a "monopoly", but if you look at the market of all interactive computing devices, Microsoft is just one of many companies. Of course, free markets do sometimes produce monopolies (or cartels), but those monopolies aren't stable, and they collapse the faster the more economically important that monopoly is; if you try to fix those problems with regulation, the cure is worse than the disease.

Welcome to Capitalism

[GameboyRMH]

In a dog-eat-dog world, you end up with one very fat dog. Not always one, sometimes two. Most of the world's beer is made by two megacorporations. Most of the world's cars are made by less than a dozen companies, and a few megacorps have the lion's share among them (Volkswagen Auto Group, General Motors, Toyota). Most of the world's computers are made by Foxconn. It's the same with everything. Capitalism only sort-of works with a small population and lowish amounts of automation, and with a credible communist rival to keep it in check. Outside of capitalism's narrow butterzone, it's just low competition between exploitative megacorps and runaway inequality until the system implodes.

I also hate the way tech has been going for a long time now, towards walled-garden computing, unnecessary use of centralized online systems, user privacy violation, and worker exploitation. A disgusting industry that I hate and would like to get out of now.

2008?

[glennrrr]

Crazy Libertarians caused government created entities like Fannie Mae and the FHA to get people to take out mortgages they couldn't afford to pay?

Re:Capatalism

[NostalgiaForInfinity]

There is no 'sharing' in Capatalism and no sense of 'common good', so you're always going to end up with a few players at the top.

That's the pre-Enlightenment view of economics. In fact, the reason the Wealth of Nations was such a breakthrough is because it recognized that individually selfish actions in a free market promote the "common good".

And it's the best way of promoting the "common good" that we know, because if you try to replace the free market with politics and government, you end up with not a few players at the top (where "few" in reality means tens of thousands or even millions, depending on how you count), you end up with one player, namely the government. And that one player is run by people who are just as selfish, power-hungry, and greedy as the CEOs of major companies. But unlike those CEOs, they are not constrained by market forces, and they can use violence against anyone with impunity.

Yes, free markets suck, but they suck a lot less than all the known alternatives.