What to do once you admit that decentralization never seems to work (2018) • Hacker News Board

Re: What to do once you admit that decentralization never seems to work (2018)

#101

Post by cameronbrown » Sat, Aug 31, 2019, 8:48 PM UTC

hn_throwaway_99 wrote: post #75

One thing I don't see argued very much is that lowering barriers to entry and providing more frictionless access usually inevitably leads to more consolidation, not less. When there is very open access (e.g. "Anyone with a computer can spin up a website!") it means that users will choose services that even have small improvements or advantages, until those services build up so much market share that they inevitably h…

Tried to articulate this same thought earlier. Thank you, this is far better than what I came up with.

Re: What to do once you admit that decentralization never seems to work (2018)

#102

Post by jasonhansel » Sat, Aug 31, 2019, 8:59 PM UTC

We already have a decentralized/federated social network with millions (probably billions) of users. It's called email, and it works pretty well--though not perfect, email illustrates that there's nothing intrinsically impossible about such systems.

Re: What to do once you admit that decentralization never seems to work (2018)

#103

Post by viraptor » Sat, Aug 31, 2019, 9:02 PM UTC

jasonhansel wrote: post #102

We already have a decentralized/federated social network with millions (probably billions) of users. It's called email, and it works pretty well--though not perfect, email illustrates that there's nothing intrinsically impossible about such systems.

Messaging systems (like emails) and social networks are not the same thing. Maybe mailing lists would be closer, but those are actually centralised.

Re: What to do once you admit that decentralization never seems to work (2018)

#104

Post by viraptor » Sat, Aug 31, 2019, 9:12 PM UTC

abdullahkhalids wrote: post #86

lanstin wrote: post #46

Earlier quoted context omitted.

Maybe someone can come up with a decentralized search infrastructure. Something agent based perhaps. Otherwise the search is both centralized and in the position to make money from ads.

YaCy got you covered https://en.wikipedia.org/wiki/YaCy Never tried it, so don't know how good it is.

It's works, but it's not amazing. You basically have to use your own spam solution for it, because the bare index will be always gamed in some way - that makes it hard in practice.

The scoring algorithm is also not amazing. For example searching for ycombinator gives you http://hackr.de/tags/ycombinator and some Twitter entries as top results.

Re: What to do once you admit that decentralization never seems to work (2018)

#105

Post by hn_throwaway_99 » Sat, Aug 31, 2019, 9:15 PM UTC

zeroonetwothree wrote: post #90

hn_throwaway_99 wrote: post #75

One thing I don't see argued very much is that lowering barriers to entry and providing more frictionless access usually inevitably leads to more consolidation, not less. When there is very open access (e.g. "Anyone with a computer can spin up a website!") it means that users will choose services that even have small improvements or advantages, until those services build up so much market share that they inevitably h…

The car dealer model is extremely bad for consumers. It seems weird to herald it as some kind of achievement.

I think you're missing my point. I hate dealing with car dealers, and I would much prefer a direct-from-the-manufacturer car buying experience. But that inevitably leads to a much greater consolidation and concentration of wealth, where consumer dollars get drained from most locales and end up in far fewer locales.

My point is that it's trivially easy to understand the negatives of things like car dealers and real estate agents (pretty much everyone I know hates dealing with them), but I think we're just starting to come to terms with the potential negatives of having huge, geographically broad, open markets where there are very few winners.

Re: What to do once you admit that decentralization never seems to work (2018)

#106

Post by rhizome » Sat, Aug 31, 2019, 9:38 PM UTC

carapace wrote: post #96

I've been reading Robert Anton Wilson this morning, so I'm all hopped up on conspiracy theory (what do you call it when the conspiracies are real?) Anyhow, the system is "works as intended", in re: mass automation of society. You want decentralization because you want power. Let me tell you a story about power and decentralization: Many years ago a friend of mine an I were rambling in the Mission district (of SF) lat…

>Anyhow, the system is "works as intended", in re: mass automation of society.

Wait, what is the idea that "the system has an intent" based on?

Re: What to do once you admit that decentralization never seems to work (2018)

#107

Post by CyanBird » Sat, Aug 31, 2019, 9:39 PM UTC

edoo wrote: post #97

coldtea wrote: post #93

Earlier quoted context omitted.

Enlightenment comes when you realize that being a consumer is not a 24/7 occupation. The same people who consume also need to be either small business owners, freelancers, or employees. In fact without the money from that, they couldn't be consumers in the first place. So what's bad for a consumer role can balance out or be very good if it helps keep more people making money as opposed to the "efficient single corpor…

Nearly every law in the US is a weaponized corporate action to prevent competition and secure market hold. That already happened. That is the root cause of the wealth disparity. You argue for more of the same. The government will not be able to fix problems it created. It will only make them worse.

This Friedman-like thinking and cute propaganda sound bytes just doesn't work tho

Chile during the Pinochet Regime is an excellent example of the Neoliberal system not delivering on the promises, as hn_throwaway_99 mentioned, economies of scale simply take over any benefit of low gov friction "free markets", and any small advantage over time becomes an overwhelming one. At which point it simply becomes a mathematical snowball effect, reducing or removing government barriers of entry won't make your particular gadget or doodad be able to compete in quality nor cost against one produced by a huge mega-conglomerate, if anything that mega-conglomerate will see if it can steal your idea or buy the patent (gov!) so it can improve its own. It is the same thing that happens in South Korea with the Chaebols

Anyhow, point being, the Kansas Experiment failed[1], Chilean Chicago Boys Experiment failed to deliver after ~17 years of authoritarian rule(When Pinochet left power country had a ~40% poverty rate)[2]

This leads to the simple idea that the less friction your system has to accrue wealth the higher wealth inequality will be, and then the goal should be to make it as easy as possible for everybody to make their own start ups, but to create as much friction as possible for said startups to gain too much economy of scale momentum. Where the threshold of where that line would be, be left to policy makers, and ideally something politically independent, akin to the central bank

[1] https://www.npr.org/2017/10/25/560040131/as-trump-proposes-t...

[2] https://imgur.com/BpwMlBy

Pg17 United Nations Document https://www.undp.org/content/dam/chile/docs/pobreza/undp_cl_...

Re: What to do once you admit that decentralization never seems to work (2018)

#108

Post by rhizome » Sat, Aug 31, 2019, 9:44 PM UTC

nwah1 wrote: post #20

sweeneyrod wrote: post #16

Earlier quoted context omitted.

> Email is decentralized by design, but centralized in practice. Why do you say that? I see email as an example of very successful decentralisation. GMail etc. have a large share of the market but they don't dominate it, and the different providers interoperate very well.

Most markets exhibit oligopolistic competition, and email is no exception. Google controls the most, yes, but they aren't the only ones to consider. The number of competitors who are significant players are only a handful. The top 5 providers likely account for the overwhelming majority of all email. Email being an open protocol allows interoperability and lowers barriers to entry, and all of that is good, but power…

>The top 5 providers likely account for the overwhelming majority of all email.

>Email being an open protocol allows interoperability and lowers barriers to entry, and all of that is good, but power is still concentrated into the hands of a few players nonetheless.

Power? In email? How exactly is that power exercised? What effect does 5 providers accounting for a majority of email have?

As far as I can tell, it has only caused too much HTML email.

Re: What to do once you admit that decentralization never seems to work (2018)

#109

Post by rhizome » Sat, Aug 31, 2019, 9:53 PM UTC

zcw100 wrote: post #11

I’m surprised IPFS hasn’t been mentioned. It’s a great example of compromise. IPNS tries to replace DNS but it’s slow so the compromised and allowed DNSLink. It’s a great application of decentralization. I want some bits that hash to X and I don’t care who gives them to me.

Nobody has ever been able to describe the use case for IPFS in simple terms, much less how discoverability works in that environment.

Re: What to do once you admit that decentralization never seems to work (2018)

#110

Post by rhizome » Sat, Aug 31, 2019, 9:55 PM UTC

jayd16 wrote: post #76

marknadal wrote: post #70

Idk what the author is talking about. Bittorrent was 40% of the world's traffic, more than Google's at that point! And we, my decentralized protocol, just grew from 2M to 15M monthly active users. ( https://github.com/amark/gun ) It seems to be working.

Torrent links are centralized in trackers and you'd see a single tracker site dominate if the legal atmosphere allowed for it. And honestly, it doesn't work as well. If you don't care about decentralization for decentralization sake its not ideal. Streaming websites don't use it. Netflix doesn't torrent a video to you.

Can't collections of tracker links themselves be federated?