Vitalik summarized 11 predictions from 2013 to 2017: most met expectations, but the rise of NFTs and the difficulties of sharding were unexpected

Source: Vitalik Twitter

Compiled by: Gu Yu, Chain Catcher

This is a small Twitter storm about some things I've said and written over the past decade, as well as my views on these topics today.

One

In 2013, I wrote an article about "how Bitcoin really helps Iranians and Argentinians." The core point: the main benefits of Bitcoin are its internationality and censorship resistance, rather than the "21 million supply cap." I predicted that stablecoins would thrive.

Last week, I actually went to Argentina! My judgment: largely correct. The adoption rate of cryptocurrency is high, but the adoption rate of stablecoins is also high; many businesses operate using USDT. Of course, this could change if the dollar itself starts to face more issues.

Two

This article from 2013 about the consequences of Bitcoin services becoming more "regulated."

The core argument: Bitcoin resists government not by cleverly knowing what legal category it belongs to, but through technical censorship resistance.

My view today: of course, Bitcoin's decentralization will allow it to survive in a super hostile regulatory environment, but it cannot thrive. Successful censorship resistance strategies require a combination of technical robustness and public legitimacy.

Three

My prediction in 2015 that we would have PoS and sharding. Honestly, these were very wrong and worth a laugh; I will share a screenshot of a presentation I made in 2015 so everyone can laugh more easily.

But what was my core underlying mistake? In my view, it was a deep underestimation of the complexity of software development, as well as the difference between a Python PoC and a proper production implementation. The ideas from 2014 were too complex, such as "1 2D hypercube."

Today, the Ethereum research team values simplicity more—both in the final design and in the implementation path. There is a greater appreciation for pragmatic compromises. Dankrad's new sharding design aligns very well with this spirit.

Four

Regarding this article from five years ago, I still 100% support my comment that "the cost of each transaction on the internet of money should not exceed 5 cents." This was the goal in 2017 and remains the goal today. This is precisely why we spend a lot of time researching scalability.

I should also add that the core idea of sharding remains intact.

Blockchain 1.0: Every node downloads everything, has consensus.

BitTorrent: Every node only downloads a small amount, but has no consensus.

Ideal: Efficiency similar to BitTorrent, but with consensus similar to blockchain.

Five

In 2012, I briefly defended the energy waste of PoW. Fortunately, by 2013, I was excited about proof-of-stake mechanisms as a promising alternative.

This reflects a broader intellectual evolution I experienced: from "X is something I must defend, so anything that benefits X must be correct" to "I like X, but X has flaws, and it seems Y fixes them, so I now support X+Y." Soldier mindset --> Scout mindset.

Six

A 2014 article about self-executing contracts. Basically, it tried to argue that making society more like a formal system is good, and we should be excited about it.

Through my work on collusion issues, I saw the limitations of this way of thinking. The fundamental problem with pushing everything into a formal system is that almost any formal system with more than two participants is attackable.

Seven

Before the altcoin market cooled down, I liked altcoins. Check out this article from September 2013.

Three core arguments: (i) Different chains are optimized for different goals (ii) The cost of having many chains is low (iii) Exits are needed to prevent core development teams from going wrong.

Do I still agree? The above parameters are less strong today because (i) chains are more general, (ii) applications are more complex, leading to greater bridging risks, (iii) viable experiments on L2. But even so, I think there are some things you can't do on L2, and there are different L1 spaces.

Eight

I was particularly optimistic about Bitcoin Cash because I agreed more with the big block argument in the scaling wars than the small block argument.

Today, I think BCH is mainly a failure. My main point is that communities formed around rebellion, even if they have a good reason, often struggle to endure in the long term because they value bravery over capability and unite in resistance rather than a consistent forward path.

Nine

These posts from 2016-17 were basically advocating for someone to build Uniswap. Clearly proud of this.

Interestingly, I got the concept of "doing something very simple and stupid, even if it's suboptimal" here, but it took me a long time to reach proof-of-stake mechanisms and sharding design.

Ten

Applications envisioned in the Ethereum white paper: ERC20 style tokens, algorithmic stablecoins, domain name systems (like ENS), decentralized file storage and computation, DAOs, wallets with withdrawal limits, oracles, prediction markets.

A lot of it is correct (basically predicting "DeFi"), although the incentive for file storage + computation hasn't taken off as much (yet?), and of course, I completely missed NFTs. I would say the biggest thing I missed in the details was the collusion issues in DAO governance.

Eleven

A more detailed view on stablecoins from 2014.

A large part of this article tried to address whether you can have a stablecoin without oracles by using blockchain data (like PoW diff) as a pseudo-price oracle.

I am now more pessimistic about this, especially because of PoS. We will need oracles. If we want stablecoins to remain robust when the dollar collapses (if that happens, by switching to their own local CPI), we will need more proactive governance.

Conclusion: My thinking about politics and large-scale human organization was more naive at the time. I was too focused on simple and complete formal models; I did not appreciate the cultural challenges and legitimacy issues.

Summary

In the early days, I did have a good intuition to avoid the craziest parts of Bitcoin extremism. There were some early mistakes, but I corrected them quickly. The mistakes of X do not mean that any specific rebellion against X will go smoothly.

Technically, I was more correct in abstract ideas than in production software development issues. Over time, I had to learn to understand the latter. I now have a deeper understanding of the need for simplicity than I imagined.