Predicting the decentralized web

I’ve experienced the beginnings of both the centralized and the decentralized web, and see many similarities between them. In this article I’ll highlight some of these similarities, and let them serve as insights for predicting 2019 for the decentralized web.

Contents

• Early adopters
• Killer apps
• The browser market
• Corporate adoption
• Tooling & development
• Predictions for 2019

Time span

I’ve focused on 1992–1995 for the centralised web and 2015–2018 for the decentralized web.

Early adopters

Up until 1994 the web was a place for students and hobbyists. This is because universities were one of the few institutions able to offer internet connectivity. There are only a handful of websites still alive from that era, with the most prominent being IMDB, Wired and Amazon. If we follow the theory of diffusion of innovations, the web back in ’94 was in its early adoption phase, and its usage grew based on interest instead of profitability.

The barrier to entry was high, since you needed the following things to get online:

• A personal computer
• A modem
• An internet service provider
• Internet connection software, such as Trumpet Winsock

By 1995 about 14 million Americans had internet access at home, despite the internet being slow, unintuitive and expensive. Reports suggest that the average connection speed in the OECD area was 56 kbps in 1995.

Even though the web itself was new, users were still aware of how slow it was, as they often had experienced much faster network connections at their universities.

Using the World Wide Web at 28.8 Kbps (or less) can be a frustrating experience: a multimedia page that takes a fraction of a second to download at Ethernet speeds takes many seconds at modem rates.

The above quote is interesting as the decentralized web faces the same challenges. People are aware that decentralized platforms, such as Ethereum, are slow and immature for consumer adoption, yet they keep on building decentralized apps. But another quote from the same era, 1995, actually turns the scalability issue into something positive.

The early Internet was developed to meet a specific goal: the interconnection of academic sites for the purpose of open scholarly research and education. In meeting this goal, the Internet was not just successful, it was too successful. The rapid growth of the Internet into sectors of the economy that it was never designed to serve (such as banks and on-line information services) has revealed some gaps in capability that were not important to early users, but are very important to the new users.

We can draw the same conclusion about the decentralized web. It’s been too successful when apps such as CryptoKitties cause network congestions and Vitalik Buterin states that developers trying to build a decentralized Uber or Lyft at the moment are screwed. Adoption of the decentralized web is outpacing its underlying blockchain technology, just like back in 1995.

Another parallel between the centralized and decentralized web is the skepticism it was met with. I am often asked “What business problem does the decentralized web solve?”. Well the centralized web didn’t solve many business problems back in 1995 either. It was more a set of promises and hype fuelled by very smart people. But still it managed to grow into mass consumption and become a public utility.

Killer apps

By mid-1994 there were 2,738 websites, and the internet was slow, insecure and too simple for use cases such as commerce. This lead it to initially only serving the early adopters. The first killer apps were within search, namely Yahoo, AltaVista and Lycos. They’d let internet enthusiasts find more websites. To an outside person this was pretty pointless as they didn’t know what to search for. The local pizza place didn’t have a website to find their opening hours, you couldn’t order anything online and streaming was still 15 years away. Yet the internet grew, as the number of websites went from 2,738 to 10,000 during 1994.

You can draw parallels to the decentralized web where the first killer apps are only serving the enthusiasts. By mid 2018 there are 1510 decentralized apps and in my opinion the only ones currently providing value are the decentralized exchanges. This renders the decentralized web rather useless to people not holding crypto. Yet the decentralized web is growing, with the following statistics:

• 104 dapps released in 2015
• 164 dapps released in 2016
• 618 dapps released in 2017
• 623 dapps released in the first six months of 2018

It’s not quite the explosive growth as in the centralized web back in the 90’s. This might be attributed to decentralized apps being business critical software as they handle other people’s money, something early websites never did. It also leads to the development process of decentralized apps being much more intricate than building static websites, which the web mostly consisted of in the 90's.

Another factor leading to the decentralized web not growing as fast as the centralized web is the ICO craze and the money that came with it. The first real killer dapp was the ERC20 token, so people spent more time trying to raise ICO money for products that have a 1–2 year time to market, than focusing on building smaller dapps.

The browser market

Internet Explorer 1.0 was released as part of an add-on package for Windows 95 in 1995 called Microsoft Plus! It was along with Netscape Navigator one of the first corporate attempts of dominating and monetising the nascent world wide web. Comparing the 90’s browser market to the dapp browser market, we see more actors, with browsers such as Toshi, Metamask and Trust Wallet gaining momentum. The difference is that these browsers aren’t plays from large organisations, perhaps apart from the Cipher browser which was acquired by Coinbase.

Corporate adoption

In terms of web presence, large corporations such as Apple (1996), McDonalds (1996), and Coca Cola (1995) were very slow to get online. In general big brands didn’t have the web on their radar as they weren’t tech companies and lacked competent IT organisations.

Currently there are no decentralized apps from large corporations, and frankly I believe it doesn’t make sense for them to try out the decentralized web other than for PR purposes. But there are hints at brands such as Coca Cola testing out blockchain technology, and considering that large corporations now have greater IT capabilities than in the 90’s, there’s a lower barrier for them to enter the space.

Tooling & development

Tooling wise the 90’s saw products such as Macromedia Shockwave, Microsoft FrontPage and Macromedia Dreamweaver. These tools were a double edged sword, as they simplified creating websites, but also lead to the fragmented and proprietary internet of the early 2000s.

It’s not far fetched to draw the same parallels to services that generate smart contract code, such as TokenGen or EtherParty. They might have their place initially in a market were competent smart contract developers are scarce, but in the long run these tools tend to get outcompeted by the influx of new developers.

Development wise in 1995 - CGI, JavaScript and HTML was used to create websites. CGI itself was rather slow and JavaScript and HTML were new standards. Browser vendors such as Microsoft and Netscape went on to implement different subsets of HTML features themselves as the interest in the web outpaced its governance organisation - the IETF. Later on platforms such as Java Swing would also arrive, born out of capitalisation attempts of the web and frustrations with its usability.

We see the same outcome here as Solidity (the most popular programming language for the decentralized web) is being challenged by platforms such as NeoVM, EOS and Lisk who all are using other languages.

Predictions for the decentralized web in 2019

Here I’ll give my predictions on what is to come for the decentralized web. The analysis I’m making is based on qualitative research, with one foot in how history has played out, and the other in 10+years of personal experience in the field of information technology.

Early adopters - predictions for 2019

The number of dapps will continue to increase and that’s almost exclusively on the Ethereum platform. Ethereum has first mover advantage, as well as being open source – which historically are strong indicators for success. It’s also driven by a welcoming and helpful community that further strengthens the ecosystem around it.

For the decentralized web to gain even more momentum, big players such as Coinbase will have to enter the market. I believe they will interweave asset management and dapps, creating a seamless experience between your wallets and the decentralized web. Coinbase will be a key player in driving the adoption of dapps to the early majority, probably sometime late 2019.

Killer apps - predictions for 2019

Decentralized exchanges will continue to gain ground, but we‘ll also see gambling become more popular. Gambling for the centralized web grew steadily 1998 and onward, but since money is such an integrated part of blockchain, I believe this will take place much faster for the decentralized web. I also take into account that blockchain scalability issues won’t be solved until 2019 – making gambling, a computationally less demanding market, ripe for growing larger.

The browser market - predictions for 2019

Coinbase will focus on one of their browsers, Toshi or Cipher, and grab most of the mobile market. They have the brand recognition among the early adopters and will aim for the early majority. They also have a fleet of great UX designers outnumbering any smaller players trying to get into this space. Another prediction is that they will launch or acquire a dapp browser for desktop computers. Coinbase is the Microsoft of the 90’s internet, having both the ambition and resources to bring dapps to the masses.

Corporate adoption - predictions for 2019

Not much will happen here as large organisations are traditionally afraid of cutting edge technology. Perhaps we’ll see something done for PR purposes, but other than that don’t hold your breath for a McDonalds or Apple dapp anytime soon.

Tooling & development — predictions for 2019

The tooling will become more competent, with frameworks such as Truffle, IDE’s such as Superblocks and more outsourced smart contract auditing taking place. Solidity will still be a poor language, but it will be the most popular one (just like JavaScript in the 90's). The collective consciousness will be smarter though, having the DAO and Parity wallet hacks in recent memory. So developers will code more cautiously and have more tools to fall back on. 2019 continues to see Solidity as the de facto decentralizd web programming language, and Truffles as its framework.

Conclusion

History tends to repeat itself and that’s what I’ve based most of my predictions on. The main resemblance between the 90’s internet and the decentralized web is how slow, expensive and insecure it all was. Yet we couldn’t help ourselves building upon it. And who knows, the decentralized web might all be a pipe dream which never comes true. But 2018 is the first year I have the same gut feeling about something as I had about the internet back in 1995.