Vitalik’s Latest Reflection: Ethereum Is Trapped in Path Dependence—It’s Time to Reboot from First Principles
TechFlow Selected TechFlow Selected
Vitalik’s Latest Reflection: Ethereum Is Trapped in Path Dependence—It’s Time to Reboot from First Principles
This article represents the most thought-provoking internal inquiry from the Ethereum community: If you were building it today for the first time, what would you build?
Navigating Web3 tides with focused insightsTranslation: TechFlow
TechFlow Introduction: On March 6, Vitalik published another long-form essay, directly calling out a long-standing “path dependency” mindset within the Ethereum community: the tendency to pursue incremental improvements atop the existing ecosystem rather than reimagining application-layer possibilities from first principles. He urged Ethereum participants to “take off their suits and ties” and adopt a bolder, more open-minded approach toward rethinking DeFi, decentralized social media, identity, and even the intersection of AI and privacy.
Full Text:
I believe we in the Ethereum world should adopt a bolder, more open mindset across many dimensions—especially regarding the application layer and how we perceive our place in the world.
The core properties—Censorship resistance, Openness, Privacy, and Security (CROPS)—must remain non-negotiable. We must not conflate “open-mindedness” with leaving people uncertain about which security guarantees the L1 will retain a year from now. We must not ask ourselves questions like, “Do we really need light clients to verify chain correctness trustlessly?” But especially at the application layer—and at the interface between Ethereum and the outside world—we should be far more willing to rethink concepts fundamentally and step outside our comfort zones.
This includes technical questions—for example: “What if AI development essentially renders browser extensions and mobile wallets obsolete within a year?”
A recent example of such a mindset shift was treating privacy as a first-order priority on par with other security properties. That implies an entirely different Ethereum application stack, since the entire stack to date has not been built around privacy. Fine—then let’s build an entirely different Ethereum application stack!
Another example this year is the growing work on network-layer privacy, both within the Ethereum Foundation and externally.
It also includes application-layer questions—for instance: “What if most of the rest of DeFi were simply a general-purpose futures market built atop high-quality decentralized oracles, enabling users to self-organize on top of it?” And: “What if the ideal decentralized oracle were merely SNARK-based verification of zk-TLS data from several mainstream news sites, performed by an M-of-N ensemble of small LLMs?”
(By the way, this intersects with AI: One consequence of AI is that it shifts “applications” from discrete behavioral categories with distinct UIs into a continuous space. Hence, the paradigm of “building fewer applications and relying on users to self-organize on top of them” will inevitably expand as a dominant pattern.)
Another example this year is rethinking the role of L2s from scratch—what kind of L2 would truly maximize synergy and mutual benefit with Ethereum?
This also extends to cultural dimensions. It’s precisely why I, @AyaMiyagotchi, and others have deeply engaged with the “milady” phenomenon. Yes, it’s a silly, cute meme. Yes, some milady adherents’ political views embarrass me—and sometimes verge on outright sycophancy (though others hold exactly the opposite stance). Yet its underlying, meta-message is clear: “Take off your suit and tie.” If you’re wearing a suit and tie, grab the nearest glass of wine and pour it over your tie—forcing yourself to rip it off and reclaim full bodily flexibility and freedom. Next time you’re invited to a formal dinner among the wealthy, actually imagine doing just that. Write down the assumption “I am a respectable person” on a piece of paper, crumple it up, and burn it. This psychological ritual unlocks an intellectual one—liberating greater creativity and expanding the boundaries of possibility.
For far too long, our Ethereum algorithm has been: “We have this existing ecosystem; what’s the next logical step to make it just a little better?” Now, our algorithm should be: “We have this exceptional L1—which will only grow more exceptional—and an increasingly rich toolkit, both built internally and imported from outside. So, given everything we know and have today, what is the most valuable thing we could do?”
If you were writing the application chapter of the 2014 Ethereum whitepaper—approaching DeFi, decentralized social media, identity, and related domains from first principles—what would you write? At least take this first step: zero out all path-dependent concerns. Pretend Ethereum’s current usage is zero. Imagine you’re the first person proposing or building an application on it—see what emerges. Even if you’re actively building today’s applications, do this anyway.
This is how Ethereum becomes powerful again.
Top Comments:
@dcposch:
Zeroing out path dependency for things like “wallets”—100% agree. Usage is effectively zero anyway, and many entrenched product decisions are simply wrong.
Zeroing out path dependency for off-chain finance—100% disagree.
The “monetary signal” capable of delivering liberating value to 100 million users must involve seamless integration with fiat rails and issuer assets—that’s indisputable. It needn’t stop there, nor be locked into those elements. It can be open-source and meaningfully superior in privacy, accessibility, and censorship resistance. But it must respect this specific path dependency—otherwise, we’re just stuck on a geeky island, entertaining ourselves.
@vitalikburterin:
Yes, I fully agree that integrating mainstream-user-facing payment apps/wallets/agents with traditional finance—enabling deposits and withdrawals—is entirely reasonable.
The path dependency I hope we shed is primarily that of Ethereum’s own application/wallet ecosystem history.
Just one random example: If you’re building a payment app, consider exploring designs that never expose 0x addresses (e.g., using one-time addresses only for deposits—and still allowing withdrawals to others’ addresses), with all payments settled internally via Railgun or Aztec.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
