What is Web3?

You might have heard a lot of talk about “Web 3.0” and “Web3 services” or the like while following cryptocurrency news. It sounds really good, like the future, but what does it mean? This is a tough one, because its not universally agreed upon but that’s ok. In this post I’m going to go over the history of “Web 1.0” and “Web 2.0” and discuss a little what a new generation of web technologies will look like, and why I disagree with the consensus in our little crypto bubble.

Recently there was a thread on Web3.0 and NFTs by Chris Dixon, called “Why Web3 Matters” where he gets wrong what the origins of the web were, and why people were so excited about Web 2.0 when it came. I am not sure if he just young and doesn’t remember, or if he is misrepresenting the evolution of the web, but here is his position and I will disprove him below. For the record he is a General Partner at a16z, and his opinion on this both matters because of his position, and is completely biased by his investment positions.

In his tweet he positions that Web1 was about “open protocols” and Web2 was about “centralized services”. This is wildly incorrect, almost completely backwards. The original Web, “Web 1.0”, was the static web. A service ran a web server, people visited the web server and downloaded information. Most services were centralized during this time, especially at the start.

Web 2.0 was the dynamic, user-generated web. No longer was the web a static unchanging web page, it was something that could be modified by users. It was the comment section. It was social media, and YouTube. It was the Ajax framework that let you customize the site to how you wanted. It was about users contributing to the web experience of other users, and themselves. This is really easy, and everyone over 35 should remember this.

So in a nutshell, Web1.0 was about users as ‘consumers of content’ and Web2.0 was about users as ‘creators of content’. It has nothing to do with the centralization or decentralization of the services offered. If anything, Web1 was the centralized web and Web2 was the decentralization of that. So Chris Dixon gets that completely wrong. And because of that, I don’t think its possible he gets the definition of Web3 right by adding together his other two wrong definitions. I don’t want to go into the rest of his thread, because I agree with some and disagree with others (no, you cannot buy “a piece of the internet” unless you are talking about investing in fiber optic cables, or Cisco).

So that brings us to what Web 3.0 really is. In the cryptocurrency community its basically been established that “web 3.0” is a decentralized web enabled by blockchain services, similar to what Chris is claiming. I’m not so sure. None of the big web players have adopted this ideology yet, and no one quite knows for sure what that will look like in practice. Will it involve Distributed Identity? If it will, then its got some time to go since that’s still a theory with some pretty major issues to solve. Basically the promise of a blockchain enabled Web 3.0 is just that – a promise.

Blockchain is also not the only possibility for what will eventually be considered the foundation of Web 3.0. It’s not even the most likely by my best guess, although I’m sure blockchain technologies, and especially NFTs, will fall into this bucket as well. What I actually think Web 3.0 will be is the AI generated web. If Web 1.0 was about static content, and Web 2.0 was about user-generated dynamic content, then it logically follows that Web 3.0 will be about dynamic content created custom for the user without their involvement. We are already seeing that, and have for years, in social media and search results and it follows that as AI improves that will spill out to the rest of internet services on the web. Soon you will go to a website and it will re-arrange itself based on what it thinks your preferences will be, and organize content based on how it thinks you will want to consume it. We see that with social feeds already, and sales platforms like Amazon, but I expect it will be true of sites as small as The Chia Plot soon. Just like the comment section that required improvements in every day technology stacks before they were accessible to the every day webmaster, AI tools will soon be ubiquitous while building sites.

AI has been improving which such ferocity that I don’t think any web service will be free of it in 5 years. Already a lot of the major web companies, like CloudFlare and Google, are adopting AI tools for network security, image scanning, data analysis, and my personal favorite – upscaling. And if you believe the vendors, most of them with actual market solutions not just promises, there is a lot more coming. And soon. And that’s why I think the real Web 3.0 will be AI driven. It might use blockchains. It might incorporate distributed identity. But it will definitely be run by the machines.

I don’t think this debate will be settled by this little Web 2.0 blog, but if you want to help keep it well out of the Web 1.0 era then please leave a comment and let me know what you think. I think the blockchain proponents may have gotten to the term first, but that doesn’t mean they will determine what the rest of the world considers to be Web 3.0. And until there is some killer app that blockchain brings to compare to some of the marvels being accelerated by AI my money is on the machines.