This is what the architects of Ethereum actually told investors they were building.
These quotes are from primary sources: whitepapers, founder podcasts, ethereum.org documentation, and the books that defined the movement. They are not paraphrases. They are the words used to raise the capital. Specific products were promised. Specific industries were named.
"You can have permissionless, decentralized applications that could support finance, social media, ride sharing, governing organizations, crowdfunding, potentially create an entire alternative web."
From finance (banking, payments, crowdfunding) to sharing economies (Uber and AirBnB-like platforms) to communications (social networks, email) to reputation systems (credit rating, seller rating on e-commerce sites) to governance, the possibilities are endless.
Whereas most technologies tend to automate workers on the periphery doing menial tasks, blockchains automate away the center. Instead of putting the taxi driver out of a job, blockchain puts Uber out of a job.
As we move into the future, we find increasing need for a zero-trust interaction system… entrusting our information to organizations in general is a fundamentally broken model.
The idea it sparked, to build a machine that could "decentralize" anything a computer, corporation or government could do, wresting power away from financial incumbents and building more enlightened social networks than Facebook and Twitter. It stuck.
Web1 was read-only, Web2 is read-write, Web3 will be read-write-own… ownership gets distributed amongst its builders and users.
Blockchains and the software movement around them, typically called crypto or web3, provide the only plausible path to sustaining the original vision of the internet as an open platform.
Millions of transactions per second, zero fees… a blockchain supercomputer making decentralized applications a reality.
The decentralized web was imagined as three connected parts: Ethereum for computing, Whisper for messaging, and Swarm for storage.
Eleven years after launch. Promise by promise.
A thesis can be evaluated. Each named use case was specific enough to be checked. Below: the original promise, the 2025 status, and the verdict.
The 11-year scoreboard
Every comparable technology of the era shipped its original promise well within this window.
A thesis is fairly judged against the diffusion curve of its predecessors. The question is not whether 11 years is "enough time"; the question is what 11 years bought elsewhere. Each technology below is matched with the actual founding promise from its inventors and the verifiable point of mainstream delivery.
The pattern is unmistakable. Every other major consumer technology of the past forty years delivered the core of its founding promise within seven to twelve years. The web. Smartphones. Cloud. EVs. AI. Web3 has had longer than any of them. What it has produced is overwhelmingly speculative infrastructure for trading the assets it issued to fund its own development.
The most damning data point isn't the delay. It's that the builders pivoted away from the original vision.
A 10-year delay on a hard problem is forgivable if the people who took the money are still working on the problem. They are not. Below: the documented pivots, in their own words, by the same people who raised the capital.
We tried for 4.5 years to put social first, but it didn't work. Dan Romero, Farcaster co-founder · January 2025 · $180M raised
Even among the most successful projects in the entire ecosystem, the real-utility rate is single digits.
Industry-wide failure statistics are sometimes dismissed as bias against a long tail of failed projects. So we audited the cohort that should make the strongest case for web3: the top 100 cryptocurrencies by market capitalization as of May 3, 2026, the projects with the most capital, attention, developer activity, and survival time. The bar was deliberately narrow: (a) demonstrably used in the real world for its stated purpose by people other than traders speculating on the token, and (b) performing that task better than the centralized alternative it set out to replace.
projects passed both criteria.
Ninety-one did not.
Cohort: CMC top 100 by mcap · Snapshot: May 3, 2026
The nine projects that passed both criteria.
In ranked order by current market capitalization. Each of these is genuinely used and arguably better than the centralized alternative, though as the right column shows, several deliver only narrow utility, and most exist primarily to service the speculation economy itself.
The ninety-one that did not, by category.
The composition of the failures is revealing. The 91 cluster into a small number of categories, most of which are pure speculation vehicles or duplicates of the same product.
A more aggressive filter, eliminating projects that pass only because they support the speculation economy itself and removing Bitcoin (not a "web3" product by its own founder's definition), would reduce the count of genuine consumer-facing decentralized successes from the top 100 to roughly two: stablecoins (a centralized product) and Monero (a privacy product).
A 91% failure rate in a curated cohort of survivors is not the long tail. It is the head.
How much is left to build? Almost all of it.
Of the eight specific use cases the founders of Ethereum named in 2014 as the reason for the project's existence: finance, social, ride-sharing, AirBnB-style sharing economy, communications, reputation, governance, and an alternative web. Exactly one has shipped a product with meaningful real-world utility: stablecoins. And stablecoins are USD-backed IOUs issued by two regulated US-facing companies (Tether, Circle), which is the most centralized product imaginable, distributed on the most decentralized substrate imaginable. It proves the rails. It does not prove the thesis.
The audit of the top 100 cryptocurrencies in § 05 makes this concrete from another angle. Nine of one hundred. And of those nine, one (Bitcoin) was never web3 by its own founder's definition, two (USDT, USDC) are centralized, and several (Uniswap, Aave, Hyperliquid, Chainlink) exist primarily to service the speculation economy itself. The genuine consumer-facing decentralized successes among the top 100 reduce, on the strictest reading, to roughly two. One of those is a privacy product.
The other seven of the original 2014 use cases are either missing entirely (decentralized Uber, decentralized email, decentralized AirBnB), tried and abandoned by the people who raised money to build them (Whisper, Farcaster's social pivot, EOS's high-performance chain ambition), or captured by speculation rather than utility (DeFi as leveraged trading, NFTs as collectibles).
The honest answer to "is anyone even still working on this?" is: a small minority. The dominant narrative inside crypto has shifted three times in five years: from "alternative web" to NFTs to DeFi to stablecoins to "AI × crypto." Each shift has been accompanied by founders pivoting away from the original 2014 thesis rather than toward it. The infrastructure has improved. The user count has not. The applications people use are still trading and storing dollars on rails that were supposed to host an entire alternative internet.
For comparison, the World Wide Web went from invention to global utility in 11 years. Smartphones went from launch to reshaping every consumer industry in 7. Cloud computing did the same in 7. ChatGPT did it in 2. Web3 has had longer than any of them, and the most-funded attempt at its single most evangelized use case, decentralized social, was publicly abandoned by its own founder in January 2025 with the words "it didn't work."
The thesis was specific. The window has closed. The deliverables are missing. And, by their own admission, most of the people who took the money are no longer trying to build the thing they sold.