Web 1.0's idealism was focused on liberal, unilateral sharing of information. People just put things up without any real commercialisation, until the dotcom era of the very late 90s. Not all the internet was the web: there was a decentralised censorship-resistant messageboard system known as USENET, and people tended to access email with local clients.
Web 2.0 was the commercialisation and polishing of the above; everything became a web app. The downside of this was everything being infested by advertising.
Web 3.0 likes to claim it's a return to 1.0 ideals of information-sharing, but if it's built on cryptocurrency it will inherit its flaws: appalling waste of energy, fundamental capacity limitations, and a tendency to collapse in fraud.
There are flaws in every evolution of the web, but none that can't be mitigated and presumably the good that each evolution brings to the world far outweighs the bad.
Jaron Lanier often makes the point that the goals of free information didn't turn out quite the way they intended
> The internet was built on a socialist model that everything should be free and accessible to all. But it also celebrated visionary tech entrepreneurs who made it big with their world-changing ideas. “How do you celebrate entrepreneurship when everything is free?” he mused. Tech companies looked to advertising to fund their operations, and that’s where all the corruption began.
Ads powered by algorithms that perform behavior-modification on it's subjects is the only model left when all that "information" is free.
> A monetized information economy will create a strong middle class out of information sharing—and a strong middle class must be able to outspend the elite tip of an economy for democracy to endure,”
Certainly, but we know what the likely alternatives were at the time: telco walled gardens (Minitel) or dialup services (AOL).
I maintain that building the internet without an integrated billing system was its most impressively radical choice. If a telco had been involved they'd have built the billing first then worked out what limited range of services they felt like selling, on which they would try to keep as much of the margin as possible rather than letting third parties profit.
The economy of monetized information would be much smaller, since you would need to "buy in"; no Wikipedia, no Stackexchange, no free courses, no free software, no free maps.
> A monetized information economy will create a strong middle class
This very much depends on whether information is a form of labour or a form of capital. What's the information equivalent of "r > g"?
That looks like a bit of a strawman. It wasn't that stuff on Web 1.0 was meant to be 'free' per a 'socialist model'. It was just sharing information and was completely orthogonal to concepts of money or value.
When you sit down next to someone on a train and start a conversation you don't start by saying 'the following exchange of words is free of cost'. Cost doesn't even come into it. It's not socialist, capitalist or pineapple. It's just words.
The signifier "socialist" in jaron's quote is almost completely superfluous, and can easily be dropped without loosing anything from his main argument, so I wouldn't fixate on that.
The point is that if "information is free", then it severely restricts the kinds of business models startups and entrepreneurs need to survive.
> When you sit down next to someone on a train and start a conversation
I don't think Jaron is trying to say you need to pay someone to have a conversation on a train. That is more like a strawman.
The content that you provide on a social media platform, which lives on a central, proprietary, for-profit company's server, is a bit different.
He begins by claiming to have coined the term "Web 3.0". Reminds me of when some other guy recently claimed to have coined the phrase, "priming the pump".
When I see a statement like this, one which seems clearly absurd, I have a hard time taking whatever follows seriously.
Well, he definitely "coined" it as in "transform it into coins". Crypto-coins, to be exact, mostly of his own currency, which has seen a quite exorbitant rise in valuation, not in the smallest part based on exuberant hyping of this "Web 3.0", of which everyone really needed to be a part (by buying Ether) or they would miss out on the next big thing.
Here's hoping Web 3.0 goes back to minimalism. Unless you're building a fully interactive web application, let's get back to server side rendering, optimized images, and minimal web requests. Now that everyone is getting faster and faster internet speeds, the web has turned into the old sloppy coding practices, where CPU speeds were doubling year after year. Just give me the information we need without all of the magical fluff around it.
I honestly don't think I would want to keep working with webdev if I had to go back to relying on server side rendering for everything, the workflow is just _so_ much worse (to me).
I'd like to hear more about your experience. I was in a technical presentation yesterday in which a web developer working on a Rails codebase that has been transitioning from backbone to React over the past couple of years talked about her experience trying to add a single page to the app. In the end, it required 14 steps, and I couldn't help but wonder whether all the scaffolding, code generation, and wiring to enable both dynamic and static rendering to work correctly (static for search engines) really provides a better experience for anyone. I get why companies like Google and Facebook developed these frameworks, but for a site of moderate size (say, a million monthly uniques), is it really a win? I look at sites like https://pinboard.in/ and feel like they really hit the sweet spot on utility, user experience, performance, and development workflow. But most of my experience is on the backend. What am I missing?
Not that I give any of these two any chance of succeeding, but the fediverse seems to me like it has much higher chances of short term success than Web 3.0.
Small (and a pretty ignorant) comparison:
Fediverse: Fair amount of clients, some users, utilizes existing technologies and creates decentralized versions of already existing paradigms based on the W3C recommendations.
Web 3.0: Uses blockchain, can't access it from your browser, has barely any users but apparently a lot of clients, inherits probably the worst aspect of .onion services (inability to just type in the URL).
However, looking purely at the financial level, Web 3.0 seems to be winning. It's where money is funneled through, because there's a small chance that it will offer something grandiose in return. Meanwhile, fediverse operates on pennies donated by the users to approachable sysadmins with some spare resources.
One of them can afford a summit in San Francisco (sponsored by a few foundations), and the second one will probably never have a summit anywhere.
You could still put all domains of an alternate domain system behind a X.web3.org that loads the whole system client side, and then connects to peers via WebRTC to load the domain X via webtorrents or ipfs or zeronet or whatever. That way it's (mostly) backwards compatible to normal websites and you can remove the "root domain" trust point later.
There were hardly a dozen alt-coins before Ethereum. I mentioned specifically Ethereum because it was Ethereum that introduced a Turing complete virtual machine that allowed people to execute code in Blockchain. Before Ethereum came into the picture people used Blockchain only for storing value. So Ethereum was a game changer and it is only 3 years old.
Ethereum developers are currently working hard on making Ethereum more scalable. It cant definitely power the web anytime soon but its getting better and better. Read about Sharding, Plasma, Raiden etc if you are interested.
There were hundreds of alt coins before ETH, they were bitcoin copypastes and no one cared.
When you start talking about offchain ETH solutions or using multiple ETH systems to scale, you are removing the only thing that provides value from Blockchain - Trust.
When an ETH sidechain can get 51% attacked or the entire system is verified through a few centralized locations, you remove the only reason you should use blockchain.
I used to drink the ETH koolaid, I now realize blockchain shouldnt be used for apps.
I used to be much more positive about decentralization. I even wrote this little decentralized side project[0] that did well on HN a few years ago.
But lately, I've become more cynical about it. I fear it just doesn't map that well to the real world. From an abstract/logical architecture it makes a ton of sense, and would be beyond amazing to have a decentralized, unstoppable, computing platform that anyone can hop on, like Ethereum promises.
But there's real computation happening in physical machines somewhere, real fiber connections between servers, real electricity costs, real national borders, real laws, etc.
These days I'm much more enthused, I think, by federation, as that seems to map more cleanly onto the real world, putting the ownership and costs and responsibilities on those who control the hardware. I was extremely excited, for example, by sandstorm.io, but it seems like it's winding down lately. I haven't really seen an equivalent that's caught my eye, but maybe I'm missing some.
I really do support the "Web 3.0" vision as defined here, and best of luck to them. If they can get that "killer app" that makes everyone want to join, and provide a little defense against detractors trying to use physical means to shut down the logical decentralization, then it would be a revolution for sure. But it looks hard to get there from here, to me.
> [federation] seems to map more cleanly onto the real world, putting the ownership and costs and responsibilities on those who control the hardware.
This is both the biggest strength of federation and its biggest weakness. How do those who control the hardware profit from federating open and free services? Federation will be a contender when (if?) somebody figures that out.
I'm old enough to remember when Web 3.0 was another name for The Semantic Web[1]. I'm sure Gavin Wood is a clever person but the opening of this article made him come off as a dotcom boom era huckster.
I experimented with Beaker Browser[1] and explored some dat:// pages. It reminded me when I first used the Tor Browser. Most of the pages looked like they were Geocities pages, and the same goes for dat:// pages - they are like something from the late nineties web. So much for the term 'Web 3.0'
I love the geocities feel :-). I'm not expecting Web 3.0 to look any different from the Web we know today. Instead it will be a Web that respects user's privacy and at the same time a Web that is more cost efficient for large companies who want cut their hosting costs and stop forking over heaps of cash to their competitors (Amazon, Google).
Yeah but I think it's the type of users creating the webpages. They sloppily throw together some HTML & CSS and it leads to this dated Geocities look and feel. (No jQuery powered single page application which hogs memory and heats up your CPU)
Maybe I'm looking at this blockchain thing all wrong.
Perhaps it's a lesser evil, whereby the foolish money is vacuumed up and held in a relatively innocuous form before it is dispersed all around where it would cause greater mischief and misery.
If we get fully cryptographic computation (I don’t know if I’m using the term correctly, but basically no one knows what you’re doing with their cycles), why would we need decentralization?
Couldn’t Isis run their irc server on AWS if it was impossible to know who is doing what?
Computation is just a small part of the system. Payment is a big one - your account can still be flagged and taken down if they ban your credit card or equivalent. Another is public services: if you're serving a publicly accessible site, for example, at some point the data must be encrypted to be shown to the user. If they can trace that back to the AWS machine, the account can still be taken down.
>[...] We see wealth, power and influence placed in the hands of the greedy, the megalomaniacs, or the plain malicious. [...] the same old dynamics.
Insanity is running the same program over and over again and expecting different results. You can tweak the configs a bit and change the running environment; but until we start editing the sourcecode (our DNA) you may expect the crooked wood of humankind to always act like humans always have.
>It was over four years ago that I coined the term “Web 3.0.” Back then, it was clear to me: Ethereum—the platform I helped found—would allow people to interact in mutually beneficial ways without anyone needing to trust each other.
Well, I'd rather we have a web of people trusting each other...
Web 1.0's idealism was focused on liberal, unilateral sharing of information. People just put things up without any real commercialisation, until the dotcom era of the very late 90s. Not all the internet was the web: there was a decentralised censorship-resistant messageboard system known as USENET, and people tended to access email with local clients.
Web 2.0 was the commercialisation and polishing of the above; everything became a web app. The downside of this was everything being infested by advertising.
Web 3.0 likes to claim it's a return to 1.0 ideals of information-sharing, but if it's built on cryptocurrency it will inherit its flaws: appalling waste of energy, fundamental capacity limitations, and a tendency to collapse in fraud.