Hacker News new | past | comments | ask | show | jobs | submit login
The data economy demands a new approach to antitrust rules (economist.com)
152 points by ravij465 on Oct 29, 2017 | hide | past | web | favorite | 86 comments



Paraphrase: "These large corporations do not look or act as trusts, so we need to make new antitrust rules to target them."

If it doesn't quack like a duck, and a duck gun doesn't work against it, maybe it's not a duck?

If it's an issue with data collection or privacy, why even talk about antitrust? If you break Google up into five smaller companies, you don't actually shrink the amount of information collected, you actually just grew it 5x as much!

Why not just pass laws specific about the data collection if that's actually the issue?


How so? Google has a 90+ percent market share. Facebook also is in a similar boat. Amazon is every increasing its market share as well. They are doing this not by competing but by leveraging user data to create good enough products. Then use their market share to integrate these products into their existing product line thus resulting in death of startups they are targeting.

Here are some examples:

- Facebook uses its data on social activity to identify trending startups. They even go further and have purchased a VPN provider to spy on traffic so they can identify startups they want to acquire or crush.

- Google uses search data to figure out what types of websites are trending and then simply creates cards that appear on top of their search results so that users never have to leave Google. A lot of websites have lost revenue or simply folded due to the loss of traffic.

- Amazon is no different. The use internal search data to figure out what products are trending and then simply creates Amazon Basic products to kill them.

The issue is they leverage user supplied data and their existing market share to create noncompetitive environments for new incumbents. It's a massive intelligence gathering operation that gives them unfair advantage by allowing them to surveil the vast marketplace and target trending startups in near realtime. Heh, the startups don't even know they are being targeted until it is too late.

They've basically become malicious actors.


US antitrust law is written to protect availability of consumer choice, not necessarily businesses from each other.

As someone who enjoys Google Cards and Amazon Basic (Facebook I will give you), I would not be happy if the government deemed the product I enjoy to be illegal under the basis of protecting my available choices.


Err, you are proving my point. You protect the availability of consumer choice by protecting business from actors who are leveraging their marketshare and engaging in uncompetitive practices. In this case, other businesses (i.e. Google, Facebook, et all)

I am with you. You should have choice. That is exactly what antitrust law is written for.


But... Wait... Are Google Cards (just the example we chose) an uncompetitive practice (as you outlined above) and be broken up, or a consumer choice that I have a right to? You are saying that they are both.


Google cards are an example of anti-competitive behavior. They lift content from sites, often without their permission, and display that data on the site. With their 90 percent marketshare, this behavior kills websites. There have been numerous lawsuits on this. The end result, is you have a web with less choice and innovation.[1] The most outrageous thing is they take this content and display ads right next to it and profit off it. All without giving a dime to the website owners.

Cards are one example, the best example is Google's plan to kill comparison shopping sites because it was eating into their marketshare. Washington Post put it best:

> Quoting internal Google documents and emails, the report shows that the company created a list of rival comparison shopping sites that it would artificially lower in the general search results, even though tests showed that Google users “liked the quality of the [rival] sites” and gave negative feedback on the proposed changes.

Google reworked its search algorithm at least four times, the documents show, and altered its established rating criteria before the proposed changes received “slightly positive” user feedback. Internal Google documents predicted that the proposed changes would reduce rivals’ user traffic up to 20 percent and subsequently reported producing the desired results once the changes were implemented.[2]

[1] https://theoutline.com/post/1399/how-google-ate-celebritynet... [2] https://www.washingtonpost.com/opinions/you-should-be-outrag...

The list goes on and on...


> They lift content from sites, often without their permission...

To be fair on this point, if the salient content is facts (for example, the expected weight of a healthy cocker spaniel), then the originating site didn't "own" the content anyway.

They certainly have copyright on the presentation of the facts, but that just obligates Google to run the original content through some sort of algorithm (or something) to make sure they're not violating copyright.


The issue is in how the laws are defined, not in any moral concepts of fairness. In court they can always claim that the consumer doesn't have to use their service, they have choice to use a competitor. The problem is that in the internet age, it is not in controlling supply but in effectively unlimited demand created by combination of zero marginal cost to add customers and network effects of the internet.


Who forces us to use Facebook and Google? Am no fan of either company to say the least, but no one I know is compelled to use them.


you have bundling issues try getting rid of youtube, if you like android. Also, monopolies usually increase prices which in this case are the advertising costs that are passed onto the consumer via more ads/ad space on google/facebook and higher ecommerce prices.


I don't think that's the most generous paraphrasing. I think the framing the article is putting forward (right or wrong) is:

The "first generation" of US antitrust rules were built around emergent giant of those times. Oil and it's transportation. Shiping. Infrastructure. Big, megalithic chunks of a maturing manufacturing economy, especially bottlenecks.

New era of commerce, new "trusts." New concentrations of influence that change the rules of the game in a corrupting way.

IDK if I'd frame it exactly the same way. Personally, I don't think we did the best job on the last generation of trusts. Data is referred to a bit to vaguely for my taste. I agree with you that working with a concept (trusts) that's nt ideal in the first place. But it's not a nonsense argument.


We can't let a few companies own our data. It's feudalism. We need data democracy.


Counterpoint: I do own my data, including my data about the people who visited my website.

Any attempt to claim "ownership" of data by the sole virtue of it collecting information "about you" would necessarily constitute a restriction of the rights of others. And it would fail utterly when you have data that is inherently about multiple people and not about any one of them individually. And assuming you could license data about you (because if you can't then fundamental parts of current society and its structures wouldn't function) then this would change absolutely nothing, as any interaction would come with such a license.

We should, by all means, have strong privacy protections, both technological and (in some very careful cases) legislative. However, framing that as "our data" would lead to unintended consequences. Let's find a framing that works.


>Counterpoint: I do own my data, including my data about the people who visited my website.

You're not actually refuting his point about this being data feudalism. What you're saying is that you're the lord of this website and in exchange for letting serfs onto it you have a right to their data. It's the same thing as traditional land/labor feudalism with just a few words exchanged.


The rest of my post refutes that point; the first line was intended to demonstrate how terminology like possessiveness over data is both ambiguous and unhelpful.

And as for the "feudalism" analogy: no, I'm most assuredly not saying that, don't put words in my mouth. If people voluntarily visit my site and send me data, the natural consequence is that I now have that data and can do whatever I want with it. Anything other than that would be an artificial restriction, and one not justified by simple possessiveness on data given that the same possessiveness applies in the other direction as well.


the natural consequence is that I now have that data and can do whatever I want with it.

You're aware that the EU quite violently disagrees with this assessment and member states, starting in 2018, will have the ability to levy cripling fines on such behavior.


I'm aware. I'm sure that'll do wonders in encouraging more incredible and compelling web applications to be created, outside the EU. It'll hurt multinational companies that have offices in the EU, make it harder for people trying to start or develop such sites in the EU, and otherwise continue to be irrelevant for the huge number of sites over which the EU has no jurisdiction.


>If people voluntarily visit my site and send me data

Plenty of serfs were voluntary as well. If your only problem with feudalism is that some serfs weren't then maybe you are a feudalist.

>the natural consequence is that I now have that data and can do whatever I want with it. Anything other than that would be an artificial restriction

No no no. Nothing here is "natural" OR "artificial." Property relations, your ownership of a website at all, the fact that you receive any data about who is visiting it, all of these things can be said to be as natural or artifical as the person speaking wishes them to be.


My problem with the term "feudalism" is that it utterly fails as an analogy. People use websites, and draw value from them, and give data to them (and random partner sites and ad networks, which is a bigger problem). They don't swear fealty to them. It's an intentionally provocative term with little bearing on concrete details, and I have about as much interest in arguing about whether it applies as I do in arguing whether data is blue, yellow, or green.

As for natural and artificial, my point is that regardless of the nature of the relationship, it's artificial to hand someone a pile of data but then arbitrarily restrict what they can do with that data. That's true whether you're talking about copyright, DRM, or personal data.


>As for natural and artificial, my point is that regardless of the nature of the relationship, it's artificial to hand someone a pile of data but then arbitrarily restrict what they can do with that data. That's true whether you're talking about copyright, DRM, or personal data.

Your ownership of a website is just as artificial as a government restricting what you can do with it. That's why I say the terms natural and artifical have no place here.

>People use websites, and draw value from them, and give data to them

Serfs use tenures and draw value from them and give loyalty to them.

The problem in your sentence is that I believe it should be rephrased like this: "People use websites, and draw value from them, and give data to their owners" which makes the feudal analogy a lot clearer and represents your own views better, for example the "I" in this sentence: "I do own my data, including my data about the people who visited my website."

>They don't swear fealty to them.

By fealty you mean what? Swearing service to a lord in exchange for a plot of land? But you expect visitors to your website to swear some of their data to you in exchange for the info on your site.

>It's an intentionally provocative term with little bearing on concrete details

You and I disagree on that then.


> Your ownership of a website is just as artificial

When you bought and paid for the server, and you're the only one who has administrative access to it, that's not at all artificial. That's a simple question of who created it and who controls it.

The natural state of affairs is for anyone who has bits to be able to process and copy those bits. Any restrictions on that, whether in the form of copyright or data-collection laws, are necessarily artificial. That doesn't make those restrictions wrong; for instance, I think it's entirely appropriate and likely uncontroversial to prohibit any public disclosure of people's physical location (e.g. home address or current GPS location), for safety reasons. I'm simply arguing that any such restrictions are artificially constructed, so we should talk about which restrictions are necessary and reasonable.

By contrast, a framing that treats data about someone as somehow "belonging" to them implies restrictions on basic things like the ability to communicate with others. (At the most basic level, you'd be talking about telling another person some personal information about you and expecting your "please don't share that information" to have some force of law; that's a restriction on that person's natural ability to communicate any knowledge they have.) Such a framing flips the polarity around from "which restrictions are necessary to protect people" to "which limited rights are granted over that data in order to accomplish things", and I think that's a fundamentally broken approach.

> Serfs use tenures and draw value from them and give loyalty to them.

Castles and towns in the middle ages got rather poor Internet service, and changing loyalties tended to be frowned upon, to say the least. Your continuous attempts to reframe things in those terms are not welcome. What changes, precisely, do you want to see? If you want to talk about regulations on the collection and use of data, let's talk about those. I'm not going to spend time trying to map that into some arbitrary and intentionally provocative analogy about "serfs", any more than I'm going to attempt to map websites to analogous models of cars or gothic architectural styles.

> But you expect visitors to your website to swear some of their data to you in exchange for the info on your site.

I'm ignoring the charged term "swear", here. I expect that they voluntarily provide data they're willing to provide in exchange for things they see value in that use that data, yes. For instance, they might provide information on what things they like or dislike, or make connections with other users, or at a lower level, their browser may make a series of requests with a given order, timing, and parameters.


>I'm not going to spend time trying to map that into some arbitrary and intentionally provocative analogy about "serfs"

Someone posted about how our current state is tech feudalism, you commented to refute that. You obviously did want to talk about this analogy.

>When you bought and paid for the server, and you're the only one who has administrative access to it, that's not at all artificial. That's a simple question of who created it and who controls it.

>The natural state of affairs is for anyone who has bits to be able to process and copy those bits. Any restrictions on that, whether in the form of copyright or data-collection laws, are necessarily artificial. That doesn't make those restrictions wrong; for instance, I think it's entirely appropriate and likely uncontroversial to prohibit any public disclosure of people's physical location (e.g. home address or current GPS location), for safety reasons. I'm simply arguing that any such restrictions are artificially constructed, so we should talk about which restrictions are necessary and reasonable.

What prevents your website from being my website? What prevents me from just walking in and taking that server? Then I could take it home, set it up and run "your" website but now it'd be mine. Well property laws prevent that and property laws are just as artificially constructed as copyright laws or data-collection laws or any other law really. That's why I'm asserting that your "ownership" of a website isn't natural and that the laws curtailing it aren't artificial. Artificial and natural never even enter into yet you keep trying to assert that.

And that second paragraph could very easily be reworked into something like this: "The natural state of affairs is for anyone who has serfs to be able to order and demand fealty from those serfs."

>By contrast, a framing that treats data about someone as somehow "belonging" to them implies restrictions on basic things like the ability to communicate with others. (At the most basic level, you'd be talking about telling another person some personal information about you and expecting your "please don't share that information" to have some force of law; that's a restriction on that person's natural ability to communicate any knowledge they have.)

The point of law is to restrict people's ability to do things. Does a framing that treats websites as somehow "belonging" to someone imply restrictions? Of course it does! Yet you're all for that curtailing of freedom.

>Such a framing flips the polarity around from "which restrictions are necessary to protect people" to "which limited rights are granted over that data in order to accomplish things", and I think that's a fundamentally broken approach.

Does a framing that treats websites as someone "belonging" to someone also flip this polarity?

>I'm ignoring the charged term "swear", here.

Then let's also remove the charged term "natural."

>Castles and towns in the middle ages got rather poor Internet service, and changing loyalties tended to be frowned upon, to say the least.

Okay, well on the web when I use your website but then decide I want to change loyalties how do I get you to give me and then delete my data? Going beyond you, how do I get anyone to do that? It seems that websites are determined to hold my data whether I want out or not, much like the serf in our example.

This is actually a perfect segway into what kind of change I'd like. Originally I had been a bit stumped about this but upon rereading your sentence above was really perfect. It seems that you and I agree that a good way to prevent tech feudalism is to give people the ability to change loyalties. In the case of user data that would seem to be some right to user data that allows me to collect it from you and request that you scrub it. Otherwise I'm not really able to change loyalties am I?


I commented to respond to phrases like "our data", which itself is a framing.

> Then let's also remove the charged term "natural."

We can call it the default state of affairs, if you like. Because there is a default, here.

Forget computers for a moment. If you tell me the statement "foo is bar", then the default state of affairs is that I now have heard from you that "foo is bar", and I can tell other people "foo is bar" or "X said 'foo is bar'". That's what I'm referring to as "natural" or "default".

There's an obvious extrapolation from that to facts like "X visited my site on this date" or "X follows Y".

Why don't we talk about policies and not analogies, here, since policies are both more concrete?

> Okay, well on the web when I use your website but then decide I want to change loyalties how do I get you to give me and then delete my data? Going beyond you, how do I get anyone to do that?

You ask nicely for them to provide such functionality, and then go "oh well" if they say no. And then you leave anyway, because unlike any strained analogy made in this thread, you're still free to leave, and you won't wind up killed or imprisoned for doing so. You're free to make use of any site you like, and stop using it whenever you like. And they're free to provide or not-provide any services they'd like.

> It seems that websites are determined to hold my data whether I want out or not

Try making that observation without using the phrase "my data" to describe "data about you" (or perhaps more generally "data that in some way mentions you"; it's not clear what you intend here, which is all the more reason to clarify). You might find that the resulting observation provides less support for your assertion. For instance, I could phrase an equivalent statement as "Websites are determined to maintain the integrity and completeness of their stored data regardless of unauthorized user actions." Sounds equally reasonable, and no less valid. The framing "my data" is presuming the point you have yet to prove.

> It seems that you and I agree that a good way to prevent tech feudalism

Still rejecting this framing device, and not agreeing to it as a premise.

> is to give people the ability to change loyalties.

I would say, more precisely, that people are free to use or stop using any site they like.

> In the case of user data that would seem to be some right to user data that allows me to collect it from you and request that you scrub it. Otherwise I'm not really able to change loyalties am I?

Of course you are; you can stop using the site at any time. You cannot, on the other hand, un-say what you've already said. While you may ask that someone pretend they've never heard it, I don't believe you should have the ability to demand that. You said it, someone observed and noted that you said it, and now you want their record of that observation destroyed.

(To be clear, here, I'm a huge supporter of privacy, and of services that provide such privacy. And I intentionally choose to use such services and avoid many others. I'd love to see more sites and applications designed to give users more control and agency over data they provide, and over how that data is stored and processed. I'd also love to see more distributed, federated services.)


>Forget computers for a moment. If you tell me the statement "foo is bar", then the default state of affairs is that I now have heard from you that "foo is bar", and I can tell other people "foo is bar" or "X said 'foo is bar'". That's what I'm referring to as "natural" or "default".

Sure but if you learn a dirty secret about me then the default state of affairs is that I can kill you or steal your server. A law which prevents that murder or theft is "artificial." Your position as I understand it is that you have a legal right to that data AND legal protection from "natural" modes of recourse. Which puts our disagreement into context – one "artifical" state of affairs versus a different but also "artificial" state of affairs.

>Why don't we talk about policies and not analogies, here, since policies are both more concrete?

Because this is a thread about an analogy - how our relationship with online data mimics the relationship serfs had with the land they worked aka data feudalism.

>Still rejecting this framing device, and not agreeing to it as a premise.

Then I'm not sure what you're doing here. The discussion is explicitly about framing our relationship with our data as feudalism. Nothing you've said so far has disproved the framing. To you my labor isn't "my labor", it's "labor provided by me" and anything which restricts the feudal lord from extracting or using that labor in any way is "artificial" because the "natural" state of labor relations is feudalism. When this discussion started I was really interested in seeing what your personal problems are with feudalism. I thought maybe it'd be something about base human rights or the power relation a lord has over a serf but it seems your main problem is that serfs couldn't switch between lords, not that lords existed at all.

---------------------------

Non-feudalism discussion:

>Try making that observation without using the phrase "my data" to describe "data about you" (or perhaps more generally "data that in some way mentions you"; it's not clear what you intend here, which is all the more reason to clarify). You might find that the resulting observation provides less support for your assertion.

Well no, I do see it as my data. If I am in front of you in line at a bank and while signing a form you glimpse my credit card number is it now not "my credit card number?" Of course not. What if your coworker figures out your HN password? Is it no longer "your" HN password? (In the ownership sense, not in the sense that you'd obviously change it.)

> The framing "my data" is presuming the point you have yet to prove.

Prove it? There's nothing to prove, you and I have a philosophical disagreement about the nature of ownership and data.

>Of course you are; you can stop using the site at any time. You cannot, on the other hand, un-say what you've already said. While you may ask that someone pretend they've never heard it, I don't believe you should have the ability to demand that. You said it, someone observed and noted that you said it, and now you want their record of that observation destroyed.

Do you feel similarly about HIPAA? Should a doctor be able to take anything you've said to them and have ultimate power over what they do with that information? What about things you mention to your lawyer or psychiatrist?


> Sure but if you learn a dirty secret about me then the default state of affairs is that I can kill you or steal your server. A law which prevents that murder or theft is "artificial." Your position as I understand it is that you have a legal right to that data AND legal protection from "natural" modes of recourse.

Well, yeah, I don't want to unravel the fundamental foundations of civil society, and I don't think you do either. Absolutely nothing we're talking about here is anywhere close to that important, and I think it's reasonable to take those as fundamental premises. So yes, I think I have a right to not be killed and to not have my equipment stolen, and I think suggesting otherwise is ridiculously hyperbolic. That's not "putting our disagreement in context", that's you saying "we have laws about murder so why not laws about how data about people can be used", which as a comparison is ridiculous on its face.

> Because this is a thread about an analogy - how our relationship with online data mimics the relationship serfs had with the land they worked aka data feudalism.

I responded to a comment that started with the sentence "We can't let a few companies own our data.". I responded, specifically, to the framing as "our data", and offered a counterpoint for why that framing begs the question. I have no interest in trying to draw analogies to the dark ages or to any other political models, when it's far more straightforward to talk about data, websites, potential policies, potential legal restrictions, and potential outcomes.

> Then I'm not sure what you're doing here. The discussion is explicitly about framing our relationship with our data as feudalism.

No, this thread is about data and the uses and control of it. See above.

> To you my labor

Nope, I said absolutely nothing about labor. Where did you even get that from? If you're going to make up positions and then argue against them, you're just talking to yourself. And one of many problems with making up analogies is that you might then be inclined to make invalid assumptions about other unmentioned aspects of those analogies associated with any given position.

> When this discussion started I was really interested in seeing what your personal problems are with feudalism.

Off the top of my head? People regularly died of numerous causes that are almost unheard-of today, had very few rights of their own, had a drastically shorter expected lifespan, had essentially no prospects for improving their lot in life, had minimal access to basic information or even basic literacy, and a hundred other problems, quite aside from likely not being able to leave and seek whatever meager prospects might exist elsewhere. None of which have anything to do with people visiting websites, interacting with web applications, and voluntarily providing data. So no, I'm not interested in talking about the state of society in the dark ages.

> Non-feudalism discussion:

Thank you for this.

> Well no, I do see it as my data.

If you're walking down the street wearing a blue shirt, and someone is sitting on the street corner keeping a tally of shirt colors, is that "your data"? I don't see any reasonable basis on which you could claim that. More importantly, it's not a framing that provides a helpful basis on which to draw conclusions.

Now, on the flip side, to provide a point on the other end of the spectrum, I do think it's dangerous and unreasonable to allow people to run facial recognition on cameras pointed at street corner and post timestamped location information of everyone they recognize. I have no problem with prohibiting that (given appropriate care to define the scope of the prohibition), and despite that I still wouldn't frame that as "because it's the data of the people walking by". I'd frame it as "allowing the publication of that information would do great harm, restricting it is a relatively small imposition, and such a restriction (assuming careful scoping) doesn't stop people from doing more reasonable things in the process".

> Prove it? There's nothing to prove, you and I have a philosophical disagreement about the nature of ownership and data.

If that philosophy leads to conclusions about what actions we should take regarding the use and control of data, then those conclusions require supporting arguments based in premises we can agree on. Right now, you're pre-supposing that "data about you" is something you should have control over, which I would counter by saying that "data in my head, or recorded by me" is not generally something you should have control over. If you want to draw conclusions based on that, they'll all be conditional on that premise, which you have yet to provide any arguments to support.

> If I am in front of you in line at a bank and while signing a form you glimpse my credit card number

I think we should redesign credit cards to operate using asymmetric keys rather than passing around an "open secret" that anyone can use to issue charges. But apart from that, I don't think it's useful to argue about whether it's "your data" or "data about you"; I think it's useful to talk about what can be done with that information. For instance, if someone spots that number and proceeds to use it to attempt to make charges, I'd call that some sort of theft or fraud.

> What if your coworker figures out your HN password?

Then I have bigger problems. I'd appreciate it if they told me first, so I could fix the gap in my security that allowed them to get it. And in certain circumstances (with more high-value accounts), one could argue that publishing that information before I've had the opportunity to change it could do irreparable harm (though any potential restriction on that would need exceptional care to avoid restricting legitimate security research). On the other hand, sites like "haveibeenpwned" keep huge databases of information about people, and use that information for the legitimate purpose of helping people. And if the compromised password ends up in a database of "known/common passwords", I don't see any reason I or anyone else should have any claim over that; it's a simple fact.


that's not really true. kids under 13 have laws around that data and how you collect it in the US. And if you can have laws to protect young kids you can have laws around what you do with older people's data.


That's a ridiculous set of laws that causes many problems for various website owners, and many very reasonably take the approach of "if you ever give us any indication that you're someone to whom these laws apply we'll ban you until you're not", rather than attempting to meet those restrictions.


> my website.

Anything publicly accessible is not yours entirely.


That first line was intentionally using the same ambiguous terminology as the post I replied to, to demonstrate that it applies just as well in the other direction.

To be much more precise: the server and the code running on it are under my control, and data is voluntarily sent to that server.


In the same sense your data isn't your data.

JoshTriplett owns his website by the virtue of control he has - it's up to him if and what is displayed on it. OTOH, data is just facts. JoshTriplett doesn't control the fact that his website displayed this-and-that content on 2017-10-29 11:11:00 UTC, nor do you control the fact that you had this-and-that official name, SSN, and bank account number on 2017-10-29 11:11:00 UTC. Those are just facts.

And as much as I understand the concern about privacy, I fear some proposed solutions just go straight up against the fundamental nature of reality. Facts are, you can't own them.


If two people communicate through a service (in whatever way), then that service falls under the telecommunications act, [1], meaning that the service should be open to external parties to allow for competition.

Quote:

According to the FCC, the goal of the law was to "let anyone enter any communications business -- to let any communications business compete in any market against any other."

It's that simple, and a shame that the law is not applied properly.

[1] https://en.wikipedia.org/wiki/Telecommunications_Act_of_1996


Laws are not applied based on its goals, but on its text. And it's incorrect to say that any service that falls under the telecommunications act must "be open to external parties to allow for competition." The Act distinguishes different kinds of services and only applies certain restrictions to some kinds.


> Why not just pass laws

Because "The Economist" specifically wants to muddle this issue and present it as merely a matter of economic configuration or transitional structural phenomena. Some of happen to think it is more fundamentally about Human Rights, such as 'Privacy of Thought', etc.

Your point re. x5 is correct. Same thing happened to John D. R. who got richer after Standard Oil was broken up.


Instead of breaking them up or enforcing arbitrary separations (say between OS and apps), regulators could force them to provide APIs for competitors to use.

For instance, Google has ~90% market share in Europe and many websites restrict access to Googlebot, which makes it impossible to compete with them on a level playing field.

Regulators could require them to offer a search API for a reasonable price, similar to the FRAND obligations put on essential patents.


The other day, Troy Hunt found a leak from a coupon website that was storing customer CVVs. If a doctor, lawyer, accountant, etc was similarly negligent they could be financially ruined (at least here in the US).

The industry may not face anti-trust issues as much as civil liability issues.


> Far from gouging consumers, many of their services are free (users pay, in effect, by handing over yet more data)

It's not free. Users pay with their privacy and most of people are ignorant how widespread their data collection is, and how much private data is collected. The companies should compete on this characteristic, but competition is limited.


It's also paid for, among other costs and externalities, by advertising. That is a $600 billion industry, online being about $100-125 billion. Assuming that's largely paid by the world's 1 billion wealthiest people (roughLY, US, EU, CA, JP, AZ+NZ), hat wworks out to a round $600/yr., or about $100-125/yr. for online services.

Contrast direct print media spend of about $150/yr.


Something about the tech crowd. So smart in general but has such a hard time with how words are used by people. Free, stealing, both words that for some reason whenever used in an article, someone feels compelled to insist on their own definition.


Free is something 'without cost'. But you do give something in exchange, your private data. So by the correct definition of free, no, Google and Facebook's services are 'not freed.

Money is not the only form of value or that can be exchanged for products and services.


We should formulate rules around data ownership - data is scarce resource that is fundamentally owned by the subjects of the data. We should probably create universal standards for data portability on platforms and mandatory licencing schemes for exclusive data or popular platforms. A youtube competitor should be able to license all the youtube videos - how would a competitor get off the ground without their catalog?


So a creator of a video should have no right to control in which sites it can be shown? Sounds like a recipe to screw creators even harder than YT already does.


If a creator uploads their video to youtube, other video sites should be able to license it from youtube. If youtube doesn't want to share the money, the creators should find another platform.


the cable industry only got off the ground and youtube only got off the ground by willfully ignoring copyright laws and engaging in a get big quick while stalling litigation strategy.

if "piracy" is the only way for superior distribution platforms to succeed then there is clearly something wrong with the laws.

i think you are probably right that requiring non-discriminatory licensing (i.e. whatever licensing deal you offer to one party you have to offer to anyone on equal terms) would be a fair and reasonable way to foster competition and innovation while reasonably compensating copyright holders.


So if I upload my videos to YouTube (because I expect to make an income from their monetization), I have to offer licenses to my videos to everyone else for free as well?


I think he's saying that you have to offer them under the same terms, so that would include the new site having to provide monetization options for your content that are similar to your existing arrangement.


Right, but for example YT won't allow you to monetize videos until you reach 10k viewers on your channel. Which means a bunch of shitty video sites can use your videos without paying you until you reach 10k viewers on that site - which may mean "never". Also, even if they do pay you, it's not like ads have guaranteed rates, it's all dynamic based on bids, which are likely much smaller in another site.


The sites would have to pay youtube for the content (ideally they'd pay something similar to how much youtube pays the creator - which would actually incentivize youtube to pay creators).

If youtube doesn't pay you, that's between you and youtube.

Maybe there should be a mechanism for creators to optout of the mutual licencing agreements but I'd make it opt-in by default.

I'd bet that the vast majority of people uploading things to youtube do so just to share their content, google shouldn't be enriched by walling off content that creators want to share with the most people.

I make a few youtube videos of my hobbies (e.g how to change the fork seals on old Honda motorcycles or pull door panels off Toyotas), I don't care if I make a dime but I'd be over the moon with joy if I got a message telling me that I inspired someone to get a bike running. Google's monopolist bologna shouldn't thwart people from getting their content out there.


Who said anything about licensing it for free? I think youtube's catalog should be available to service who can cough up as much as youtube.


I'm fascinated by what's going to happen in the post (monopolized) data economy. When data acquisition can happen autonomously, beyond HITL. Taking that data and building, growing, and strengthening intelligence could scale up something huge and interesting. We're currently building a decentralized data and machine learning marketplace and network to answer the question: What would happen if agents could sell their data in open markets?

We think it would be cool to see that data be used in turn to help continuously train ML models. It starts to look like an integral part of AI economies.


Have thought about building a similar service in the past however more focused on IOT. This is definitely a very interesting space and one of the few areas where tokens make sense as an integral part of the offering.


I agree. We originally wanted to build an appliance for it, but hardware is hard. So we abstracted out what we could and while we use "people" contributing data, the reality is that any agent (IoT, etc) on the network can participate.


You're working on Synapse right? I've been following that project. I really like the vision.


Super appreciated. Are you working in the space?


Not really but Im into crypto and this is one of the few projects I think makes sense.


> More transparency would help...

I would advocate for the ultimate transparency. The dissolution of privacy altogether.

Nothing benefits the incumbents more than the subjects being passionate about not giving the information they give them to anyone else. This has the same effects as employees not disclosing their wages: it only benefits the employer, in the game of assymetry of information. If the information is public, then the little-start-up that could would effectively be able to attack the giants.


This is a notion that sounds great at first, but it's entirely shallow.

- There are tons of information pieces that functionally need to be private, for example passwords, and tons of information pieces that better just should be kept private for your own benefit, like medical records.

- You're not helping us customers. If every company knows everything about us, then every company can exploit our mental vulnerabilities to get us to buy more things we don't need.

- Privacy is a basic human need. Lack of privacy can lead to severe mental issues, as illustrated by Panopticons.


> There are tons of information pieces that functionally need to be private, for example passwords, and tons of information pieces that better just should be kept private for your own benefit, like medical records.

Passwords are a solution of authorization , not authentication. Your face is not private but it can also be used as a password.

Medical record information is valuable only because it is private. If it were public, it wouldn't be considered as important. Much like wage data.

> You're not helping us customers. If every company knows everything about us, then every company can exploit our mental vulnerabilities to get us to buy more things we don't need.

The opposite on two accounts: today, the company that holds that data uses it against you without you knowing, precisely the requirement of privacy makes it impossible to understand how a company is using that information the way you describe. And secondly, nothing better than to erode a company's benefit of using that information than competing with others that do the same.

> Privacy is a basic human need. Lack of privacy can lead to severe mental issues, as illustrated by Panopticons.

Im not a psychology expert, but I'm willing to argue that privacy is a defense against other injustices, not valuable in itself.


> Medical record information is valuable only because it is private. If it were public, it wouldn't be considered as important. Much like wage data.

Ummm, thats not true at all. A whole lot of people would not want all their friends knowning every little detail about their medical history. There are lots of embarrassing issues that people have that nobody would want to be made public.

Think STDs or mental health issues.


The information is valuable only in so far as it helps making decisions about it: someone should know if the other person has AIDS, or Terminal cancer, for making lots of decisions. Hiding that information might help the sick person, but it doesn't give good results in the end to the system.

Another reason to defend privacy in this regard is to use it as a tool against discrimination. But its as good as thinking that the way to prevent robberies is to have no assets. It might be useful today,but its not a goal, it is a compromise.

Embarrassment is not something to indulge. It is a feeling that springs only out of society.


If your salary is public information, then any future employer can peg to it, rather than the market rate. It is to your benefit for your compensation to be privately held.


Employers know a lot more about salaries than you do, considering they have a lot more information points, reliable information data than you. The information asymmetry is too pronounced to be overcome by mere negotiation tactics.


If everyone's salary is public information, you can see what they're paying everyone else and get at least that much for the same job.


Combined with collective bargaining, not so much.

Star-system employment as well, mentioned for completeness, though that doesn't scale to populations: everybody cannot be a star.


It's not to my benefit at all. In every salary negotiation past my first job out of college I have said "I expect to be paid $X to switch jobs" and I was paid $X (or very close to; obviously, sometimes you kick a little back) to switch jobs or I didn't switch jobs.

What you do is what's important. Make a proactive argument for your value and you don't have to fall back on your current compensation.


What if every time they thought "we got him/her for a measly $X, sweet"


Just get two competing offers. Then the question flips from "how little will the candidate accept?" to "how much can we afford?".


In addition to raw data, the processed data is also valuable. It might be good for trained ML models to be available to all, for example. Those are the competitive advantage.


Trained ML models could be a reasonable trade secret, after all, its not hard for an engineer that knows about it to spill the beans naturally as he looks for another job. But an ML engineer with the data facebook has, cannot do the same somewhere else.


Like enforcement?


In the status quo, the overwhelming majority of user data lives in silos owned by the incumbents, and the incumbents lock down that user data to build nigh-impenetrable moats around their walled gardens through powerful network effects.

I've always been a huge proponent of a different model of thinking about user data and building user-facing services/applications, whereby instead of building up a centralized silo of user data pertaining each service, services request read/write access to parts of a encrypted, user-controlled data source to access/store their service-specific data.

Sophisticated users can then choose to host their own data for greater control, and others can choose to delegate the storage of their data to a trusted third-party storage provider (entrusted only with the ability to _store_ their data, as the actual data will be encrypted with their keys/passphrases).

Either way, users can rest assured knowing that they can migrate any/all of their data to another provider at any time, and that any service accessing their data through this model cannot prevent them from sharing/forking their own data to be accessed by another service competing in the same space.

Under this model, services would have to compete only by the value they bring to users, and will no longer be able to hide behind lock-in and network effects!

I truly believe this new model of data access will become a powerful enabling force for innovation on the internet, which has far too long been stifled by a focus on petty battles for market dominance through winner-take-all network-effects rather than on building the best product.

If you're interested, projects like remoteStorage [1], Hoodie [2], and Solid [3] offer building blocks we can use to build applications that operate in a model like this today (though with the caveat that you'll have to roll your own encryption layer, and there are not many storage providers to choose from). I'd love to hear about other projects like these if anyone is aware of them, and would like to invite everyone to experiment with these projects and think about how we can improve the developer and end-user experience of services that operate in this model.

Nevertheless, I realize that this will be an uphill battle for sure, because services compelling enough by their own merit to break through the deep-seated network effects of incumbents will have to come first, before this data-access model can begin to reach critical mass and foster a network-effect of its own. And that is starting to seem like an impossible battle to win, because even newcomers today have huge profit-minded incentives to build up their own silos of user data from day one, in hopes of building up network effects and securing a comfortable moat of their own one day.

I still have hope that we can arrive at this new data-access model organically one day through purely technological innovation, and have been working on a framework for building applications like this that aims to improve the developer ergonomics of this data access model through modern tooling (integrating a client-centric query & caching mechanism like GraphQL and Apollo).

But if regulation like those suggested in the article can help speed up the transition by forcing incumbents to open up their silos, I'd welcome it with open arms.

[1] https://remotestorage.io/

[2] http://hood.ie/

[3] https://solid.mit.edu/


> the incumbents lock down that user data to build nigh-impenetrable moats around their walled gardens through powerful network effects.

You say "network effect" like it's a bad thing. The internet is valuable because it lets anyone talk to anyone. The distributed web is not valuable (today) because it lets nobody talk to nobody.

"Metcalfe's law states that the value of a telecommunications network is proportional to the square of the number of connected users of the system (n^2)."

Networks are valuable because they're winner takes all. One winner-takes-all dweb is superlinearly more valuable than 50 competing dwebs. That's why messaging networks end up federated.


Even in your scenario I see it possible to collect overall user behavior characteristics and correctly classify your “anonymous” profile.

It’s the overview effect that gives incumbents nearly insurmountable advantage.


Check out what LowGlow is talking about below. Their site is at Synapse.AI


So what about the governments that use these corporate "giants" as a front for upstream data collection?

Yeah… I don't see things changing unless people start voting with their behaviors en masse.


Simply disallow startup acquisitions for companies above a certain value/marketshare. This will prevent them from killing competition. For example the social networking space would probably have been disrupted several times by now had Facebook not been allowed to purchase Instagram and Whatsapp.


Maybe?

It also likely would have reduced investment in social startups, especially those that didn’t have growth to indicate “breakout” status.


> reduced investment in social startups

Sounds like a feature, not a bug.


Not if you're trying to disrupt FB, surely?


It demands a government that can actually keep up with the pace of modern technology. Waiting 4 years for major administration changes is too long for the 21st century. Silicon Valley seems set on disrupting everything, so why not the government as well?


The reason why governments don't innovate faster or disrupt more is because they are responsible for the disrupted. A startup is not.

When a startup succeeds in disrupting an industry, they gain the benefits, and grow as a consequence. The businesses and individuals they disrupt do not have power over the startup.

When a government succeeds in 'disrupting' (perhaps 'innovating' is a better word) then they gain benefits from votes they win. Unlike the startup, they also have to deal with cost of votes they lose from the disrupted.

TL;DR: Startups take the spoils of innovation. Governments get spoils and the pain. Governments therefore appear more risk-averse.


> The reason why governments don't innovate faster or disrupt more is because they are responsible for the disrupted.

There's some truth to what you're saying, but we're all talking as if an enormous government (the national U.S. government) is the only way to attack issues. For reference, the U.S. populace is the third largest one in the world, and it's certainly a very diverse one. Being all things to all those people is a very tall order.

Keeping more issues at the state and local level allows for faster innovation, faster turnover of elected officials, smaller feedback loops, less expensive communication with elected officials, etc.

It might not be appropriate to regulate the internet at the local level, but if more were decided that way (i.e., most social issues), there would be fewer national issues left to keep track of and debate. Issues like data privacy would therefore be more prominent in national elections.

I frankly don't see how we can have such a large and detail-oriented national government long term. If China and India "crush" the U.S., it will at least partly due to the unforced error of the ever-growing scope of national U.S. politics.


As it stands western democratic governance exists in no-mans-land where everything is too little too late. This was always the case however as the pace of change has accelerated the old institutions and structures of governance are now totally unfit for their purpose. This is perhaps why the Chinese autocratic model will succeed in the 21st century and traditional western democracies will fail (NB: not a value judgement on my part, just an observation).


I don't know about that. China is just as bad as the west when it comes to privacy, and they use it to help crush dissent in their citizens.

As for the environment, while they've been taking steps to remedy that, they're still behind the West on this.

I also don't think I agree that democracies are slow to adapt. When there's the political will, democracies can get new laws passed quite quickly. The problem is when people don't care or disagree about what to do.

An "autocracy" like China avoids the problem of disagreement, but only by ignoring all disagreements and following the rulers, whether their decisions are good or bad. Usually bad, judging by the censorship, forced organ harvesting, disappearings of dissidents, etc.

It's easy to try extrapolating China's current growth and think that China will become the world superpower. Except they're still the underdogs, and a lot can still go wrong. They could plateau, like Japan did. Xi Jinping will eventually be too old to rule, and it'll be a coin toss as to whether or not his replacement will be competent. As wages in China rise, companies that outsourced to them could move to cheaper countries. If they suffer a recession, I don't know what will happen, but with a government that loves cracking down on disagreement I don't expect it to be pretty.

China's on a good trajectory, economically speaking. But it's too early to say who will win out in the 21st century.


I'm not sure what privacy has to do with any of this? You're right that nothing is a given and the tide could turn against China however there is a given precedence for stagnation due to inability to make decisions quickly or at all (as is happening in the US with the hyper-polarization of politics). This was one of the main causes of the fall of the Roman Empire.


Western democracies are "slow" because the low-hanging fruits of problems that can be solved fast has long been taken care of.

What remains are never problems with an easy fix. Almost by definition, governments are fine-tuning a system that, by and large, is already the best the world has ever seen. Any change involves trade-offs that need to be taken into account.

To use an analogy: after 20 epochs, you need to be using a drastically lower learning rate for your ML model than when you first started training, or it will fail to converge.

Progress in China isn't so much the effect of autocratic structures, but of starting from a lower baseline. It's really easy to grow the economy when you're dropped into a scenario of almost medieval infrastructure, but a working administration and enough capital. First, you build a road. Then, a school... As the Chinese economy closes the gap to the west, it will start mirroring their problems more and more, as well.


Democratic governments are slow by design. A dictatorship (or Kingdom or whatever form of "one person to rule them all") would be extremely fast, a more centralized government wouldn't be as fast but still far faster than the current decentralized versions used in most countries. That also means bad decisions could reach people far faster, in the worst case without a chance to correct them anymore.

Democracies fix this by requiring slow legislative processes. You bring in a bill, it needs supporters, it gets discussed in whatever form of working group your legislation uses, the public can discuss it, can form protest against it and so on. Sure, this takes time, but it reduces the risks of really bad things slipping through, which is far more important than "Meh! Silicon Valley just takes a week to bring a new product, I want my government MVP now!!"




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: