The case for a decentralized internet Tamas Kocsis

Three years ago,

I started building a decentralized web

because I was worried
about the future of our internet.

The current internet we are using
is about gatekeepers.

If you want to reach something on the web,

then you need to go
through multiple middlemen.

First, a domain name server,

then a server hosting company,

which usually points you to a third party,

to a web hosting service.

And this happens every time
you want to reach a website on the web.

But these gatekeepers are
vulnerable to internet attacks

and also makes the censorship
and the surveillance easier.

And the situation is getting worse.

Everything is moving to the cloud,

where the data is hosted
by giant corporations.

This move creates much,
much more powerful middlemen.

Now, move to the cloud makes sense

because this way it’s easier and cheaper

for the developers
and the service operators.

They don’t have to worry
about maintaining the physical servers.

I can’t blame them, but I found
this trend to be very dangerous,

because this way, these giant corporations

have unlimited control
over the hosting services.

And it’s very easy to abuse this power.

For example, last year, a CEO of a company

that acts as a gatekeeper
for nine million websites

decided, after some public pressure,

that one of the sites it manages,

a far right page, should be blocked.

He then sent an internal email
to his coworkers.

“This was an arbitrary decision.

I woke up this morning in a bad mood

and decided to kick them
off the Internet.”

Even he admits,

“No one should have this power.”

As a response, one of
the employees asked him,

“Is this the day the Internet dies?”

I don’t think we are actually
killing the internet,

but I do think that we are in the middle

of a kind of irresponsible
centralization process

that makes our internet more fragile.

The decentralized, people-to-people web

solves this problem
by removing the central points,

the web-hosting services.

It empowers the users

to have host sites they want to preserve.

On this network, the sites get downloaded
directly from other visitors.

This means, if you have a site
with 100 visitors,

then it’s hosted
[by] 100 computers around the world.

Basically, this is a people-powered
version of the internet.

The security of the network
is provided by public-key cryptography.

This makes sure that no one
can modify the sites

but only the real owner.

Think of it like instead of getting
electricity from big power plants,

you put solar panels on top of your house,

and if your neighbor down the street
needs some extra energy,

then they can just download
some from your house.

So by using the decentralized web,

we can help to keep content
accessible for other visitors.

And by that, it means

that we can also fight against things

that we feel are unjust,

like censorship.

In China, the internet
is tightly controlled.

They can’t criticize the government,

organize a protest,

and it’s also forbidden to post

a kind of emoticon to remember the victims
of the Tiananmen Square Massacre.

With the decentralized web,
it’s not the government that decides

what gets seen and what doesn’t.

It’s the people,

which makes the web more democratic.

But at the same time,
it’s hard to use this network

to do something that is clearly illegal

everywhere in the world,

as the users probably
don’t want to endanger themselves

hosting these kinds
of problematic content.

Another increasing threat

to internet freedom

is overregulation.

I have the impression

that our delegates

who vote on the internet regulation laws

are not fully aware of their decisions.

For example, the European Parliament
has a new law on the table,

a new copyright protection law,

that has a part called Article 13.

If it passes, it would require
every big website

to implement a filter

that automatically blocks content

based on rules controlled
by big corporations.

The original idea is
to protect copyrighted materials,

but it would endanger many other things
we do on the internet:

blogging, criticizing,
discussing, linking and sharing.

Google and YouTube
already have similar systems

and they are receiving
100,000 takedown requests every hour.

Of course, they can’t process
this amount of data by hand,

so they are using machine learning

to decide if it’s really
a copyright violation or not.

But these filters do make mistakes.

They’re removing everything
from documentation of human rights abuses,

lectures about copyrights

and search results

that point to criticism
of this new Article 13.

Beside of that, they are also
removing many other things.

And sometimes, these filters
aren’t just removing the specific content,

but it could also lead
to loss of your linked accounts:

your email address,
your documents, your photos,

or your unfinished book,

which happened
with the writer Dennis Cooper.

It’s not hard to see
how a system like this could be abused

by politicians and corporate competitors.

This Article 13, the extension of these
automated filters to the whole internet,

got strong opposition

from Wikipedia, Github,
Mozilla, and many others,

including the original founders
of the internet and the World Wide Web,

Vint Cerf and Tim Berners-Lee.

But despite this strong opposition,

on the last European Parliament vote,

two thirds of the representatives
supported this law.

The final vote will be early 2019.

The result is important,

but whatever happens,

I’m pretty sure it will be followed
by many other similar proposals

around the world.

These kinds of regulations
would be very hard to enforce

through a decentralized web,

as there is no hosting companies.

The websites are served
by the visitors themselves.

I started to build
this network three years ago.

Since then, I’ve spent thousands,
tens of thousands of hours

on the development.

Why?

Why would anyone spend thousands of hours
on something anyone can freely copy,

rename, or even sell?

Well, in my case,

one of the reasons was
to do something meaningful.

During my daily regular job
as a web developer,

I didn’t have the feeling
that I’m working on something

that had a chance to be a bigger than me.

Simply, I just wanted to make
my short presence in this world

to be meaningful.

Last year, the Great Firewall of China
started blocking this network I created.

This move officially made me the enemy

of the government-supported
internet censorship.

Since then, it’s been really
a game of cat and mouse.

They make new rules in the firewall

and I try to react to it as fast as I can

so the users can keep hosting content
and create websites

that otherwise would be censored
by the centralized Chinese internet.

My other motivation
to create this network was worry.

I fear that the future of our internet
is out of our control.

The increasing centralization
and the proposed laws

are threatening our freedom of speech

and, by that, our democracy.

So for me, building a decentralized web

means creating a safe harbor,

a space where the rules are not written
by big corporations and political parties,

but by the people.

Thank you.

(Applause)

三年前,

我开始构建一个去中心化的网络,

因为我
担心我们互联网的未来。

我们当前使用的互联网
是关于看门人的。

如果您想在网络上获得某些东西,

那么您需要
通过多个中间商。

首先是域名服务器,

然后是服务器托管公司,

它通常将您指向第三方,

指向网络托管服务。

每次
您想要访问网络上的网站时都会发生这种情况。

但是这些看门人很
容易受到互联网攻击

,也使审查
和监视更加容易。

而且情况越来越糟。

一切都在转移到云端

,数据
由大公司托管。

这一举措创造了
更强大的中间商。

现在,迁移到云是有意义的,

因为这样

对开发人员
和服务运营商来说更容易、更便宜。

他们不必
担心维护物理服务器。

我不能责怪他们,但我发现
这种趋势非常危险,

因为这样,这些大公司

可以无限
控制托管服务。

而且很容易滥用这种权力。

例如,去年,一家担任 900 万个网站守门人的公司的首席执行官在

受到公众压力后决定,应该封锁

其管理的网站之一

,即最右边的页面。

然后,他向同事发送了一封内部电子邮件

“这是一个武断的决定。

我今天早上心情不好

,决定把他们
踢出互联网。”

甚至他也承认,

“没有人应该拥有这种权力。”

作为回应,其中
一名员工问他:

“今天是互联网消亡的日子吗?”

我不认为我们实际上正在
扼杀互联网,

但我确实认为我们正

处于一种不负责任的
集中化过程中

,这使得我们的互联网更加脆弱。

分散的、人与人之间的网络

通过移除中心点

、网络托管服务来解决这个问题。

它使用户

能够拥有他们想要保留的主机站点。

在这个网络上,网站
直接从其他访问者那里下载。

这意味着,如果您的网站
有 100 名访问者,

那么它
由全球 100 台计算机托管。

基本上,这是一个以人
为本的互联网版本。

网络的安全
性由公钥密码术提供。

这样可以确保没有人
可以修改网站

,只有真正的所有者才能修改。

可以把它想象成不是
从大型发电厂获取电力,

而是将太阳能电池板放在房子的顶部

,如果街上的邻居
需要一些额外的能源,

那么他们可以
从你的房子里下载一些。

因此,通过使用去中心化网络,

我们可以帮助
其他访问者访问内容。

意味着我们也可以与

我们认为不公正的事情作斗争,

比如审查制度。

在中国,互联网
受到严格控制。

他们不能批评政府,不能

组织抗议

,也不能张贴

纪念
天安门大屠杀遇难者的表情。

对于去中心化的网络
,决定

什么可以看到什么不可以看的不是政府。

是人,

使网络更加民主。

但与此同时,
很难利用这个网络

来做一些在世界各地显然是非法的事情

因为用户可能
不想危及自己

托管这些
有问题的内容。

对互联网自由的另一个日益

严重的威胁是过度监管。

我的印象

是,我们

对互联网监管法进行投票的代表

并没有完全了解他们的决定。

例如,欧洲议会
有一项新的法律

,新的版权保护法,

其中有一个名为第 13 条的部分。

如果通过,它将要求
每个大

网站实施一个过滤

器,根据受控规则自动阻止内容

由大公司。

最初的想法
是保护受版权保护的材料,

但它会危及
我们在互联网上做的许多其他事情:

写博客、批评、
讨论、链接和分享。

Google 和 YouTube
已经拥有类似的系统

,它们
每小时都会收到 100,000 个删除请求。

当然,他们无法
手动处理如此多的数据,

因此他们正在使用机器学习

来确定这是否真的
侵犯了版权。

但是这些过滤器确实会出错。

他们
从侵犯人权的文件、

关于版权的讲座

和搜索结果

中删除了指向对
新第 13 条的批评的所有内容。

除此之外,他们还
删除了许多其他内容。

有时,这些过滤
器不仅会删除特定内容,

而且还可能
导致您的关联帐户丢失:

您的电子邮件地址、
您的文档、您的照片

或您未完成的书,


与作家丹尼斯·库珀(Dennis Cooper)发生过。

不难看出
这样的系统如何

被政客和企业竞争对手滥用。

第 13 条,将这些
自动过滤器扩展到整个互联网,

受到

了 Wikipedia、Github、
Mozilla 和许多其他人的强烈反对,

包括
互联网和万维网的原始创始人

Vint Cerf 和 Tim Berners-Lee。

但尽管遭到强烈反对,但

在上次欧洲议会投票中,

三分之二的代表
支持这项法律。

最终投票将在 2019 年初进行

。结果很重要,

但无论发生什么,

我很确定世界各地都会
出现许多其他类似的提案

由于没有托管公司,因此很难通过分散的网络执行此类法规。

这些网站
由访问者自己提供服务。 三年前

我开始建立
这个网络。

从那时起,我在开发上花费了数千、
数万小时

为什么?

为什么有人会花费数千小时
在任何人都可以自由复制、

重命名甚至出售的东西上?

好吧,就我而言,

原因之一
是做一些有意义的事情。

在我
作为 Web 开发人员的日常日常工作中,

我没有
感觉到我正在做

一些有机会比我更大的事情。

简单地说,我只是想让
我在这个世界上短暂的存在

变得有意义。

去年,中国的防火墙
开始封锁我创建的这个网络。

这一举动正式使我

成为政府支持的
互联网审查的敌人。

从那时起,这真的
是一场猫捉老鼠的游戏。

他们在防火墙中制定了新规则

,我试图尽快对其做出反应,

以便用户可以继续托管内容
并创建网站

,否则这些网站会
被集中的中国互联网审查。


创建这个网络的另一个动机是担心。

我担心我们互联网的未来
是我们无法控制的。

日益集中化
和拟议的法律

正在威胁我们的言论自由

,进而威胁我们的民主。

所以对我来说,建立一个去中心化的网络

意味着创造一个安全港,

一个规则不是
由大公司和政党制定的空间,

而是由人民制定的。

谢谢你。

(掌声)