The Future Is Decentralized | Hacker News Reader

The Future Is Decentralized

modernconsensus.com

60 pointstodsacerdoti55 comments

Can't anyone still share information using the normal internet? You don't need the blockchain in order to set up a website and start sharing information.

Few choose to do this because it is more work than using a centralized system. Using Facebook or twitter is much easier to use.

What does the blockchain really change? So far, it doesn't seem to be any easier than sharing information the traditional way.

IMO it's not more work, there just aren't teams of psychologists designing flashy signup forms and apps to make you want to do it.

I don't know if that strictly goes to what the article was talking about, but nevertheless:

Blockchains can prove that two parties once agreed on something, cheaply. That is remarkably valuable information (of such bricks economies are built). And it is truly remarkable that there is nothing in the blockchain that a government can do about it. Ditto anyone else.

I believe it is financial information and preference data that is difficult to share at the moment and that the article alludes to liberating. For example, donating to Wikileaks was extremely challenging via Paypal and other traditional money handlers. That obviously has both up and downsides.

This is new. I live in an age where my cooperation might be strictly necessary for somebody else to tax me. The practicalities haven't changed much, but the theoretical change is quite profound.

> Blockchains can prove that two parties once agreed on something, cheaply.

it's not as cheap as trusting a third party. The evidence for this is that block chain as a form of financial payment system is still way outstripped by commercial credit card providers. And it's not just existing/legacy momentum either, because if you look at china, the advent of digital payment systems on the mobile _could've_ been using the block chain, but it hasn't.

> nothing in the blockchain that a government can do about it.

And unless the entire transaction _only_ occurs on the blockchain, a gov't can always do something about it using the rubber-hose law of security.

Blockchain (sometime) can replace third party by some open source code. For example on decentralized Exchange platform you directly exchange your assets ; no trust, no third party needed. And it is cheaper than third party based exchange.

Of course, blockchain does not make sense in all the cases. But third party does represent a cost, and can sometime represent a risk (eg becoming dominant thanks to network effect, then extracting rent), or slow innovation (are you are dependent of this third party).

Please keep in mind that blockchain is not only about cryptomoney.

> ... cheaply ...

For some values of "cheaply", of course. Reliance on a mutually-trusted third party, where feasible, is a lot cheaper than any blockchain.

- often you need to rely on multiple mutually-trusted third party (eg transferring money from your bank account, to a a bank account in another continent) and that have a cost

- Blockchain sometime bring a higher level of automation, reducing cost

- And more importantly, there are many cases where the "mutually-trusted third party" becomes a "dominant power" in its area and extract rent from it. App stores come to mind, but that the case of most "platforms"... Decentralization + open source + token to incentivize the system seems to be able to have cheaper, better and more transparent stuff.

"Blockchains can prove that two parties once agreed on something, cheaply." - well actually you can do that with electronic signatures, much cheaper than with blockchains. The problem that blockchain solves is a very specific timestamping of messages between many participants. For two participants just signing a date is enough - the problem is if A singes two checks to B and C and you want to know which one was first.

1. Click != agreement. Cf: phishing.

2. https://xkcd.com/538/ - applies for any sufficiently large entities and certainly governments.

Sorry can’t edit, but a perfect exhibit for (2) is on HN right now: https://news.ycombinator.com/item?id=24545236

I agree, but I think "more profitable" is the underlying reason, not "easy."

Facebook/twitter/Whatsapp/etc. could be a lot more distributed, open, etc. It wouldn't be harder. As you said, this is what the internet and www were designed to do. You don't need blockchain, but this is just another example of good decentralization tech that exists.

It wouldn't have been hard to make whatsapp open, and/or distributed. It wouldn't have been harder to make it interoperate with other clients, and it would have been more useful that way. It just wouldn't have been worth as much money. Centralization, monopolisation and such are the value, when you go to sell your app to Facebook.

Email is distributed because it was invented before centralizing and monopolizing became so valuable. It's easy because there are lots of easy to use clients available. It's open and decentralized because it reached decisive adoption before the dotcom boom.

"Easy" to the end user is, by and large, a byproduct.

I don't think it's ingenuous to point to a html file, apache and say "there's your decentralized solution."

It isn’t so obvious to me that you can make say, WhatsApp be distributed as easily as making it centralized.

It seems like all kinds of things like identity, message routing, nat-traversal etc, are harder in the distributed case a compared with a centralized server farm.

It’s also not clear to me that email is a good example of ‘distributed’. For a start most users actually do use centralized email services like gmail because it’s easier, and secondly even in the days before that, email has always relied on DNS, which is at best only partially distributed - i.e. although the participants are in a federation distributed across the network, there are still centralized authorities governing the root servers, the issuing of IP addresses, and of domain names.

Today no one will find the information you share on your website without centralized search engines. A more decentralized internet could change that. "Blockchain" better named DLT however is not needed for that, but it may push the desire to decentralization. If people can have a "de-central bank account" that no one can delete they maybe start to think that all other online accounts you have should work that way too. No more questionable suspended twitter account and deleted YouTube channels.

The future is not decentralised because it's far too complicated.

Here's an alternative, more practical option to centralisation: let's call it the 'WordPress' model. Here's how it might work:

Hosting companies have a standard installation API for server web apps. This lets developers create server web apps that can be installed on any hosting provider that supports that installation API. A simple one-click process.

Businesses and indivduals can shop around to find a suitable hosting plan that will run the app. Like WordPress, there will be cheap-and-cheerful hosting plans at one end of the scale and at the other end are managed services with SLAs (Service Level Agreements) and premium support.

The open, standard installation API enables data to be transferred from one hosting provider to another allowing easy data transfer from one hosting provider to another. (Or at least an easy way to download your data).

The "WordPress" model is not a "pure" decentralised model but it's far more practical. It is also highly unlikely to happen because installation of server apps is currently impossible to do in an easy and simple way, particularly for non-technical users. But imagine if it was possible and the countless opportunities it would unlock for developers. It would also mean management of servers is offloaded to hosting companies while developers can simply concentrate on building their apps.

(Docker, Cloudron, Sandstorm, command-line installations etc are neither easy or simple for non-technical users no matter how much developers may insist otherwise.)

What you describe is also referred to as the self-hosted model. I agree with your assessment. Self-hosting can be a powerful model if we can make the server application model easy to use and manage, similarly to how it works on mobile phones. But the shared decentralisation model, aka blockchain, also has a role to play here. Digital identity should be sovereign and this can be modelled well using blockchains. To see more work in this direction check out https://identity.foundation.

This model is already being built: check out platforms like the internet computer & holochain for example.

Isn't this fundamentally a political issue? Even unbreakable encryption and stuff like shortwave broadcasts can be outlawed by a government which will lead to a very few number of holdouts on the run from law and the rest of society is going to comply.

True change can only happen if people passionate about this manage to convince the REST of the 95% to back them and adopt the ideals.

Once decentralization becomes more widespread, it can impact politics too. Look at "liquid democracy" initiatives in South America for example.

There needs to be a bigger benefit to the end users before decentralized services catch on. Widespread hard censorship of company run services might be just the kind of thing that decentralization needs to stop being a niche thing for people willing to put up with a diminished experience for the nebulous benefit of controlling your data and experience. Vanishingly few people want to deal with running their own services so the annoyances of not having control over what exactly happens on $socialMediumPlatform is smaller than the hassle of trying to move to mastodon then having no one to follow.

We need Justin Bieber to use mastadon.

Is he even that hot of a celebrity any more? I was not in that demographic at all but he seems to have fallen into the weird void of the 'formerly super famous' to me.

Either way it's an old strategy for businesses, pay or convince a celebrity to use your platform to bring their fans along. You want a wider span of 'influencers' though for sure because you don't want Mastodon to become the place with all the Beiber fans.

I think one of the more killer features brought to the table is using walletconnect to log into dapps. I can navigate to an app ‘log in’ using my own wallet and it allows me to purchase items without needing to create an account.

This is a breath of fresh air compared to modern ecommerce sites that ‘require’ you to send over a truckload of personal information, create a login, verify your email etc just so you can purchase a single item of clothing.

Most ecommerce sites let you purchase as guest, not requiring personal information except your credit card data.

Which includes your full name and address, right? I’m sure at that point it’s pretty trivial to link it to the data hoard dedicated trackers have on you.

How does the merchant ship your single item of clothing w/out you giving them your address?

You create another Dapp, which connects the merchant with a PO-box like warehouse which has numbered drop locations. You pay for a drop location, and the merchant gets the id of that drop location only. They send it there, the warehouse Dapp has to pay somebody to do a physical verification that a drop happened. Then you get notified of the drop, and go pick it up.

You're right, for the average user this is a much better workflow than just given a shop your address and the shop sending you a package.

When you have to buy cryptocurrencies you have to pay your country's payment system, then you have to provide information to government or a bank account. It's just an contradiction of blockchain always said its a "Privacy-Safe"thing.

I think we need decentralized hardware. Citizens should be able to setup wireless mesh networks themselves.

No it is not and it never will be. There is no business model for it. The business models are in centralization and those central points have the money to make an unbeatable user experience. Decentralized solutions can not compete and will thus never gain traction.

When we start colonizing other planets, things will have to be decentralized because latencies will become prohibitively large ...

Cryptocurrencies are a great solution for some reality. It's just not our reality now nor in the near future.

If we start colonizing planets. Anyway, even then it seems quite likely that each individual planet will still be very centralized.

Imagine being a colonist on Mars. It's not like it'll be feasible to homestead in the first few decades, if ever. You depend on centralized government for the very air you breathe. It's the very opposite of living in a decentralized society.

We will live free in the asteroid belts and Oort cloud.

Those will probably become the most militarized zones in the solar system when planetary governments realize the WMD potential of an asteroid plunging into the sun's gravity well.

I semi-seriously believe that addiction to connectivity and low ping times is the "great filter." When push comes to shove very few people would be willing to move to Mars because the internet there will be terrible.

what a lack of imagination. You could copy paste your argument changing the nouns and apply it to any open source endeavour.

Of course there is strong resistance from the powers that be, but to claim never is an incredible act of Hubris.

There is relatively very little money in open source. Also the scale of running a large application is very different compared to running an open source project most of which ironically are centralized in some way if not via GitHub, then certainly by the presence of a BDFL who drives the project.

uniswap.io

What I feel like many people miss in this argument is that there's still currently no reason for the vast majority of people to want or care about the future being decentralized. Not only does majority of the population think companies like Facebook aren't manipulating them (these people will claim they're manipulating the "other side"), but even the people that do realize that Facebook is manipulating them don't care. They find the ads useful, and enjoy the recommended videos. They'll watch The Social Dilemma and go right back to browsing Twitter.

There's still no good reason that anyone outside of a technical audience holding certain ethical beliefs would want something other than a decentralized system. Most individuals don't care about being their own bank, for example.

In my eyes, the only way the future becomes decentralized is through a massive change in the fabric that makes up much of the developed world.

There is a part of "The Future" that will always be "The Future". Maybe this is one of those parts.

>and open protocols can create a Web 3.0 free from the power of centralized mega-corporation

This repeated thinking that "open protocols" will create an escape from centralized institutions is flawed.

After reading hundreds of essays on "decentralization", most of them ignore the following fundamental economic issues:

- open protocol specifications (RFC drafts, ActivityPub whitepaper, open source implementation on Github, etc) -- all of them -- do not put money in everyone's bank account to _pay_ for running that protocol. (My previous comments about this.[0][1])

- protocols do not provide humans assistants such as system administrators or a salaried staff of IT workers to run & maintain decentralized systems for others. (E.g. my previous comment[2] that us uber geeks on HN know how to run Git but it doesn't matter because we don't have the time nor enthusiasm to run decentralized git on our home laptop, or Gitlab on Raspberry Pi to serve our repos to others. It's easier to just use centralized Github.)

If you're thinking that money and staffing are "out of scope" for a decentralized open protocol RFC, you are correct! But that's also why decentralized protocols don't lead to the decentralized utopia that enthusiasts envision. Consider that both the protocols TCPIP an HTTP are already decentralized -- and yet one decentralized node spelled "thefacebook.com" (and later changed to "facebook.com") is now feared as too centralized. Think deeply about how & why that happened. If your mental framework is limited to thinking about "protocols", Facebook being centralized is a mysterious contradiction to decentralized tcpip/http. However, if your mental model includes economic forces, there is no contradiction.

>web remains community-focused and user-oriented, just like it was originally intended to be.

Yes, Tim Berners-Lee developed the open http protocol and also open HTML language but he was originally trying to get scientists to share their data more easily. He and his colleagues' salaries and their computers and servers to run http+html were paid for by the government. His decentralized vision didn't scale out to a billions of users because most everybody else doesn't have the same financial situation. The missing discussion about economics is the same missing factor in his newer SOLID project to re-decentralize the web.[3]

If everyone keeps writing essays on "open protocols" as being the savior from centralized behemoths, we're not advancing the discussion. We're actually going backwards. Look at history to see that open protocols do not achieve that goal as intuition would lead you to believe. Instead, integrate realistic economic incentives into your predictions for a decentralized future and write new essays that take that into account. Think about why the original Bitcoin whitepaper didn't predict that mining would shift from a wide open democracy of home computers to a handful of miners in China. Even the open protocol of Bitcoin blockchain couldn't prevent economic incentives to consolidate into mining pools. A protocol can't stop a handful of miners in China to spend more money on custom ASIC chips and run thousands of them in datacenters to shift the balance of blockchain power from the hobbyists that mined on home computers with AMD graphics cards.

Maybe a thought experiment in a different domain to help break out of "open protocol" thinking: The blockbuster film industry of English-language movies is too "centralized" in Hollywood. Hmmm... so let's create a new language (maybe Esperanto?) so that it breaks the power of Hollywood and 1000 decentralized film communities making Esperanto movies can flourish in places like Alabama and Montana. Well, Esperanto or any other synthesized language will not solve the centralization problem. The English language is already decentralized. You're proposed solution is at the wrong level of abstraction. Look at the economics. (E.g. millions of dollars to pay for desirable actors and special efx is out of reach for most hobbyists regardless of the spoken language.)

[0] "openness of a protocol" idealism such the early tcpip internet leaves out the fact that governments paid the salaries of people to implement a protocol and also paid for the computers to run a protocol: https://news.ycombinator.com/item?id=20231960

[1] Signal protocol is open and free (free-as-in-speech) but there are still economic forces (not free-as-in-beer) that affect decentralized deployment : https://news.ycombinator.com/item?id=20232499

[2] decentralization works against the aggregate choices of how most people want to allocate their time & money: https://news.ycombinator.com/item?id=21535703

[3] https://www.google.com/search?q=SOLID+decentralized

I like the author’s enthusiasm, even if I don’t personally think proof of work blockchain is a general solution (proof of stake blockchain for specific applications is more interesting to me).

I went to the distributed web conference a few years ago (and I was thrilled to have good and long conversations with Tim Berners-Lee and Vint Cerf) but I am losing some interest in distributed web because other issues are more pressing that are only partly helped by distributed web: surveillance capitalism; attention economy; decoupling much of my time from using digital devices; government interference with free information flows; income inequality; lack of social discourse across political parties; etc. For me personally, these are the issues that I choose to address.

A little off topic, but I have recently updated the way I use tech: use freedom.to to make time wasting web sites and apps unavailable for large blocks of time during the day; reorganized my email and calendar to use Google’s services (I pay for ProtonMail and FastMail but I find that compromising on privacy for personal efficiency is better for me - as an author and developer I get a lot of email that I want to process in a personal way, but also be hyper efficient); I leave my iPhone at home, relying on my Apple Watch for receiving texts and calls (when I am out in the world, digital life can be too large of a distraction).

I have tried hard to use Mastodon social, but it takes too much of my time. I give myself about 45 minutes a day to be on HN and Twitter and that is sufficient to get notified of new open source projects, news, etc.

So, morally and philosophically I support people working on decentralized tech, but at the age of 69, I put my personal effort into personal productivity and enjoying life.

I hope that the web becomes more decentralized. I wonder, what catalyst would cause decentralization? My grandma doesn't care about decentralization and I am not sure she would be convinced even if I did my very best to explain it to her. I don't think it would change her habit of connecting with family via Facebook and using Google for search, mail etc.

"You can follow him on Twitter" ...oh, the irony. Some good views, buy come on, at least do what you preach and use mastadon.

The title is missing a "(2010)" note ... or was it 2000? I guess some things are always "the future".

could they do something about the snake-oil peddlers everywhere online these days? ...and blockchain/crypto won't be transformative until they ditch the deflationary obsession: it's all a Ponzi scheme! if you want people to use the "Distributed Economy", don't build tools for hoarders. ...and don't keep changing the protocol all the time(it kills adoption -- because there's no way spend it, at least not legally: no services).

> Decentralizing the web through new technologies like blockchain and decentralized protocols can ensure that the web remains community-focused and user-oriented,

Let's be realistic. A safe-place for anonymous money transactions is never ever going to be community-focused. It may start out community-focused but we all know it is going to end up like 4chan.

I think you should elaborate more on that, because your argument is really suffering currently.

4chan is a community oriented place on the Internet just like any other. Just because it doesn't cater to your tastes specifically it doesn't mean it's a negative example of a community.

When people mention "community-focused", they often omit that the community refers to their own community.

4chan is better for your mental health than facebook, twitter or istagam.

How would any of the centralized cesspools be any better than any of the other centralized cesspools?

In the end, aren't you still in a cesspool?

You're saying 4chan is healthy, perhaps because you have an affinity for it? My daughter and her friends believe tiktok is healthy, again, because they have an affinity for it. I understand that people find their communities in these centralized places. That said, having an affinity for one centralized website or another doesn't make that centralized website necessarily "healthy".

You may argue, "But Ah Haaa Bilbo0s!!! Decentralizing these sites won't make the decentralized places any less cesspool-like than the centralized versions." Perhaps not, but health and clean content are not really the point of decentralization, distributed control is the point.

We shouldn't confuse our own tribe's centralized system as the ideal of "healthy". Someone downthread here put it best:

When people mention "community-focused", they often omit that the community refers to their own community

I find it very amusing to see miffed people crawling out of the woodwork to reply to your jest about 4chan because it's a constant reminder that a huge chunk of HN is being very civil and charitable on one tab and reading a bunch of racial and homophobic slurs without batting an eye on the tab next to it. Gives a whole new meaning to civility and charitability, and indicates pretty well to whom it is directed.

> The Future is Decentralized

tHE fUTURE IS qUANTIZED

Reply on news.ycombinator.com