Tech and Antitrust | Hacker News Reader

Tech and Antitrust

stratechery.com

88 pointskaboro39 comments

Couldn’t agree more with the analysis- I just wish some of the politicians running on a “break up big tech” platform could show they’ve thought about these issues half as much as the author has. They seem to mostly be appealing to voters emotionally without ever diving into the significant details like “what is a monopoly”, “are these companies monopolies?”, “what good will come from breaking them up?”

Yes, as always, analysis of the problem is the first step, and this is where the discussion ends. The real problem is not even finding a solution, but getting it accepted. Nobody blogs about that.

Well, you can't completely understand the calls for breakup until you understand that the politicians don't really care about the traditional monopoly issues. They're upset that Facebook, Google, and Twitter have as much control over the political discourse as they do (note the missing Amazon in that list, modulo the fact that Amazon's CEO happens to own a major newspaper). If society is like a big collective brain as it thinks about the big issues, those companies have installed filters on a significant portion of the neural links and they aggressively use them.

I personally think this is a bad thing as well, and fully support breaking them up for this reason alone. In a perfect world, we'd carefully and thoughtfully craft new laws to address this matter; in reality, we're going to probably "creatively" read some existing laws in new and unpredictable ways to get some unprincipled compromise of what several groups want.

So I'm likely to be both unhappy at the theory and the practice of how they are actually broken up. Isn't government fun. But I do agree they need to be broken up.

This analysis is then useful as a conventional reading of anti-trust laws, but I fully expect that if a conventional reading of the laws won't get it done, then either new laws are going to be made or unconventional readings will be made.

I also think that these platforms are basically boned in terms of voters and public opinion. Consider two sets of forces on these platforms: First, the set of forces that want to insist that they carry any given piece of speech, and second, the forces that want to insist that they censor any given piece of speech. (Observe I'm not breaking this up on conventional political lines; all sides have their lists of both types of things they want.) I submit that there simply is no solution for a single universal platform that combines all of them. Almost all even remotely controversial speech is going to be on the mandatory list from one party and the banned list from another, which means that in the public opinion war, every political side is going to have an endless list of grievances against these platforms, literally no matter what they do. There is no solution for platforms at this scale. This is ultimately what is going to kill them; even if the government doesn't break them apart, something will.

In fact, in my first paragraph, when I said they "aggressively use" the filters, I'm not saying they even necessarily have a choice, or particularly accusing them of "incorrect" usage. I think at this point, they're simply at a size where they're boned no matter what they do with those filters. I do have my own list of particular localized grievances, but if you don't have one now, you probably will soon. The latest YouTube demonetization purge is hitting a lot of weird channels, including... a channel dedicated to making completely a-political cartoon parodies of the Final Fantasy games? Everyone's going to be pissed at something. I don't think there's ultimately a solution to "the one big site that everyone uses". Applications to the stock price of the relevant entities left as an exercise for the reader.

> They're upset that Facebook, Google, and Twitter have as much control over the political discourse as they do

This is the part that resonates with me the most. It's about power. The tech giants have amassed power in a way not seen before. It's different than Standard Oil and the Bell phone monopoly. On the balance, consumers aren't being harmed as they would in a typical antitrust scenario.

> On the balance, consumers aren't being harmed as they would in a typical antitrust scenario.

I would argue that they are. If private info is the new currency we're paying for access to these services, the price has been constantly increasing as more and more information is collected through tracking and "partnerships". This is one reason why it would be nice to put a dollar value on private data: to make this price increase obvious.

Real competition might look like delivering similar services without requiring as much tracking data to do so.

One of the problems with claiming that customers are being harmed by the privacy invasion is that the companies can demonstrate that they are only making a few bucks per year per customer on these "privacy invasions", making the most naturally legally defensible definition of the "damage" being done only a few bucks per year, per customer. The argument at that point that the customer receives far more value than that is very, very easy.

Showing that they do much, much more damage to privacy to make that several bucks is going to be legally challenging, and trying to convince a judge that the damage is yet higher because of my aforementioned nebulous concerns about "the social fabric being censored at the metaphorical-neural level" is going to fail unless you get a particularly activist judge, because they're not supposed to be ruling on the basis of such abstract concepts like that. (That's supposed to be what Congress is making laws based on.)

One of the challenges to my mind here is precisely that on the whole, individual customers aren't being harmed. On the whole, individual customers come out ahead. For most people, this is an acceptable bargain and they are substantially on the winning side of it. It is only once a single entity aggregates substantially the entire market that society and democracy starts to be substantially harmed, even as at every moment, there was never a transition point where individual customers changed to being harmed.

I think the issue of pricing is harder than that - damages and the ammount earned don't inherently line up.

If I sit at an intersection and record the cars passing and sell it to the municipal government for say $500 and I note 250 cars they didn't lose $2 each. Just because you can make money off of someone else /doesn't/ mean it is exploitation.

Conversely if someone broke a MRI Machine to steal the copper wire and sell it for $5 it doesn't mean the hospital suffered only that much in damages.

"I think the issue of pricing is harder than that - damages and the ammount earned don't inherently line up."

If you read again, you'll see I acknowledged that already. It's just that the easiest number to defend as the "damages" is the amount earned. But that number is so low that we're going to be trying to claim numbers literally 3 or 4 orders of magnitude higher. That's going to be much harder. Not necessarily impossible, but much harder.

> If private info is the new currency we're paying for access to these services

I would argue that it is not info that's the important issue here, it's share of attention, and the harm suffered by consumers is the unknown effects that is having on their personal wellbeing, as well as society overall.

Of course, there's nothing illegal about this, and forming any sort of a proof of harm would be incredibly difficult. I'm not sure what the proper path is from a legislation angle, but I believe if critical thinking was more valued and promoted, people would be much less vulnerable to harm.

More than that... they are effectively filtering what news and information you see in general terms. More so than even the mainstream media ever could. It's been pretty well established that the people in charge of these organizations have a political bias and are not only able to, but willing to actively shift the messages that people see.

The mainstream media decides what gets published before that content ever even reaches Facebook or Google. In terms of censorship power, the mainstream media is actually further upstream than social media and it can be argued it has more power to decide ultimately what gets published and what doesn’t, especially considering it purely uses human editorial judgment rather than algorithmic.

> I would argue that they are

I don't see the argument. You characterize trackable information as currency, which is a different argument. Information is traded, but is not currency. The market puts a value on data, which degrades IMMEDIATELY, until it's worthless to anyone (eg MySpace data cannot be monetized anymore).

> If society is like a big collective brain as it thinks about the big issues, those companies have installed filters on a significant portion of the neural links

This statement seems like something most people would agree with until you dig deeper.

Do you think there should be zero filters? What about threats of violence? Targeted harassment? Terrorist recruiting? Child abuse? State-sponsored propaganda?

Living animal brains continuously prune malfunctioning cells and connections to stay healthy. Is it possible that some filtering of the "big collective brain" is similarly essential to healthy functioning?

Do you have specific examples of content that is being filtered that, if it weren't filtered, would likely result in better collective decisions?

If you have specific examples, are they pervasive enough to warrant shutting down beneficial filtering?

I wasn't making normative claims; I was being descriptive.

The politicians object to the fact that so many filters are under the control of one entity, which isn't them.

I'd also point out that there was an Internet prior to Facebook, and if anything, I'd actually call it incrementally less of a cesspool. Without everybody piled into one big room it was way easier to carve out a view of the Internet that wasn't so full of concentrated crap. Lots of people keeping their own bits of the net clean to their standards worked for me.

Now I will be normative. I see the Internet as a reflection and a consequence of society. To say that you're going to, say, eliminate all child porn from the Internet is saying that you're going to eliminate all child porn from the world. I mean that very literally; no metaphor. You can't do the former with anything less than what it would take to do the latter, and there's no reason to believe that's possible.

You can't solve this problem with filters. It doesn't matter whether that truth is a good thing or not, it simply is the truth. Your text casually assumes the filters are effective, and thus offers a seemingly unresolvable dilemma about "why do you want so much child porn on the internet, jerf?" but the filters are not effective. So, given the filters can't do those things, but can do other things, it's unsurprisingly that the politicians take an interest in this.

This is really just the old question "why can't the government solve all the problems with more regulation and more people looking over more shoulders?" translated into digital terms, and the answer is, by the time the government is large enough to do that, the people the government has to be made out of have brought the problems in with them, only now they're where they can hide themselves easily.

> eliminate all child porn from the world

Somewhat reducing CP is good, even if you don't eliminate it completely.

> the filters are not effective

Using your chosen example, there is very little CP on Facebook, Instagram, YouTube, Twitter and Amazon. That's prima facie evidence that filters can be effective, even if they're less than 100% effective.

> why can't the government solve all the problems with more regulation

I'm starting to see the theme of your argument now. If X achieves less than 100% perfection, then we should eliminate X, whether X is content filtering or government regulation.

Some drivers fail to stop at red lights and stop signs, causing accidents and injuries. But few people would agree to eliminate red lights and stop signs. Even though they're flawed, they are better than nothing at busy intersections.

I don't think that breaking them up is entirely necessary. Classifying them as public spaces in private ownership and extending similar freedoms from government to these large social media companies seems like a relatively reasonable approach. As to the counter argument that they are "private companies," corporations and companies exist because of the government. The positions allow for not only group ownership, but also for such limited liability that companies can collectively kill thousands and not be killed themselves. This can and should be offset by limited capability in scenarios where a corporation may have an unnatural leverage against competition.

I tend to be very libertarian leaning, but corporations are not individuals, nor are they held to the same standards, and as such shouldn't be allowed the same freedoms. Restricting business is natural, and civil rights laws have already made business owners do what they might otherwise not want to.

Extending freedom of speech rights enforcement to organizations with over 10m members or managing over 5b/year in revenue is not unreasonable. They already have platform protections.

I agree -- in fact, I'd argue that attacking the problems of social media through the lens of antitrust will result in a far worse outcome than directly addressing the problem head on that these companies have created a privately owned, global public square. The laws need to be updated to reflect the novel realities of the emergence of these platforms, which was likely an inevitability given the arc of the development of the Internet/web, and not something our forefathers could have forecasted when writing applicable laws and regulation around speech.

As Peter Thiel wrote in Zero to One, both small and giant companies will try to creatively redefine the definition of the market they are addressing, in the first case claiming that they absolutely own a small market segment of questionable relevance and in the second case claiming that they are small fish in a much bigger market than the one where they are enjoying a monopoly. So the results will be determined not by some objective standard but by relative political power of these companies and their adversaries.

As for durability, I personally fail to see how google and facebook are not durable monopolies - the network effects are so strong. In google's case it is even more than that - it is almost impossible to have a product of comparable quality without the users and the information about their behaviour, which is impossible to obtain because everyone uses google. Gone are the days when good search rankings could be inferred from the link graph of the web.

to me, ML creates an interesting question concerning monopolies when strictly looking at data-results. If I have the best data possible, you can't build a better-competing model. However, forcibly splitting data up among competitors creates worse models, which is not best for consumers.

Apple's walled garden's the only place a pro-consumer, pro-privacy stick has been successfully used against the broader software industry. The only alternative to achieve similar levels of privacy and safety are a combination of hyper-vigilance and just not doing very much. I'd hate to see that little walled city of sanity breached until the worst behavior of the various software snoops (Google, Amazon, Facebook, but also credit card companies and others) has been legislated out of existence, assuming that ever happens.

>Apple's walled garden's the only place a pro-consumer, pro-privacy stick has been successfully used against the broader software industry

Except that Apple has given the iCloud encryption keys of millions of Chinese users to the Chinese government [1], after forcing users to move their data to the Chinese data centers [2]. This gives the Chinese government access to Apple user photos, docs, messages, and other user data at rest, demonstrating that Apple is more for-profit than for-privacy.

[1] https://www.theverge.com/2018/7/18/17587304/apple-icloud-chi...

[2] https://appleinsider.com/articles/18/02/24/apple-to-move-chi...

[1] says "the encryption keys are also stored in China, raising the possibility the Chinese government could gain access to it"; nothing even close to "Apple has given the iCloud encryption keys of millions of Chinese users to the Chinese government" (my emphasis).

> demonstrating that Apple is more for-profit than for-privacy

In China. In the other 194 countries in the world, this isn't true.

Intentionally ignoring millions of Apple customers is a No-True-Scotsman.

Fair.

Do you have any suggestions for any alternatives that are better in regard to user privacy?

From the article:

"Apple is leveraging their iOS monopoly into an adjacent market — the digital content market — and rent-seeking. Apple does nothing to increase the value of Netflix shows or Spotify music or Amazon books or any number of digital services from any number of app providers; they simply skim off 30% because they can."

If I download and use the Netflix app, Does Apple really take 30% of the monthly Netflix subscription payment? If so, that seems unreasonable because Netflix is doing most of the work of streaming the content.

It depends on whether you sign up within the app or sign up on the Netflix site and simply log into the app. Paying through Apple will give them the 30% cut.

OK. Then that's legal.

If you could only get access to it through Apple, that could be illegal. Of course, only to the extent that Apple is a monopoly. Or at least a majority market holder? So, yeah, trying to argue that in front a 9 supremes, who are sticklers for getting this sort of thing right, is not likely to go well. We really should have changed the laws, and only after the laws were changed, should we have gone after the tech companies. Would have been much more successful.

Doing what we're doing is just a display for the purposes of getting votes. It's like the politicians don't even think the issue is serious, they just want to try to get some votes.

Which, I suppose, is their only real goal anyway.

The difference being, that all app subscriptions and payments made on device are forced to use Apple payments as a supplier. The fact that Netflix is diverse enough allow an independent web payment option is a red herring, because there are many iOS apps that are not. Consumers are forced to use Apple payments by Apple.

I tend to think that Apple should keep IAPs for actual in-app features but things like ebooks and video subscriptions should be treated like physical goods, where app makers can offer their own payment (and fulfillment). Amazon doesn't pay 30% for items purchased with the Amazon app, but it would if it offered Kindle purchases. This distinction makes no sense, since ebooks aren't app features or game gems or whatever. IAPs for such features make sense, since otherwise every app would just be "free" with pay-to-unlock and Apple getting nothing for running the store.

Nutty idea for discussion. Should the US go in the other direction and officially sanction some of these monopolies with the tradeoff that they have to accept some extra oversight/regulation (ex. telecoms)?

It could be argued that some sort of oversight of these things is sorely needed. The well documented ability of mobile apps to influence behavior, and collect and exfiltrate the detailed personal data of US citizens at population scale presents an unprecedented national security risk - especially in the age of AI.

Until/unless we find some solution for regulatory capture, this direction is a lot less appealing to me. I hardly think the current regulation of telecoms is a good example of what's best for the public.

That approach is more sensible for exlusive markets like utilities. If a better product exists they can go the way of Yahoo quickly. Now it could be argued they have too much inertia and splitting wouldn't actually help much - even so it raises another big question.

What regulation would even help? What is the actual explicit problem that could be addressed and is it even something addressable constitutionally? There is plenty of sloganeering and FUD but little thought into what would help from even limited to first order effects. Break up Amazon retail into five and the one with the name would win.

Also how does that qualify as a national security risk? I know the term is used as a thought terminating cliche for "shut up and obey" but I just don't see it in any way which respects democracy. People have a right to communicate with the outside world which allows data gathering and influence. Taking action to leave people completely "uninfluenced" would be totalitarian control.

The "unprecedented threat" reminds me of the anti-Catholic "if elected they would be a puppet of the Pope" - which shows utter historic ignorance given how kings interacted with the pope. He wasn't close to being in charge in a far more homogeneous and theocratic environment.

Is all this antitrust talk directed at tech companies not just a way for media to direct that energy away from themselves?

These companies provide valuable platforms. If you want to bring competition back, you will need a way to get ubiquitous platforms without the monopolies.

The platforms are going to be provided by decentralization technologies. Such as cryptocurrency, various content-centric networking, p2p everything apps, etc. Federation, where people host open services on servers and connect them is more popular now and may be a significant factor for awhile also. I assume p2p systems will eventually be more widespread once people figure out how to do it, since p2p can scale better/more easily.

I really don't see how p2p scales better in economic terms. Because crypto currencies lack central authorities they need mechanisms that are highly energy intensive to guarantee consensus, thus driving up prices for transactions. The same thing is true for a federated social network. The sum total of all federated instances is harder to manage than facebook.

The benefit of decentralisation is redundancy and fault tolerance, not the ability to scale.

There are less energy intensive cryptocurrency scaling techniques.

true, but they'll never be as efficient as a centralised system, because intrinsic to the trustless design is a proof of work or proof of stake, which is to say the investment of resources to drive up the cost for any given actor, to make owning the system impossible. There is no way around this by design.

We have two "monopoly" problems here... the legal, and the political. Even for companies that are probably not legally monopolies, like Facebook, consumers are very uncomfortable with them, and politicians are leveraging that consumer fear. So if antitrust doesn't work, think about where politicians might turn next.

Or they may just go on making blathering promises with no intent to keep them, like they do about, say, Citizens United.

Reply on news.ycombinator.com