Best of farcaster.xyz/vitalik.eth on Kiwi News
@vitalik.eth on Farcaster · Most upvoted by Kiwi News readers
63 stories by @vitalik.eth on Farcaster have been shared on Kiwi News since 26 December 2023, with 156 upvotes and 50 comments in total. The latest was shared on 4 October 2026. The 50 most upvoted are listed below.
Text under a title is an AI summary of the linked article.
- hot take re "too much investment in infra"
- vitalik.eth: How I would do creator coins
Vitalik Buterin argues that crypto creator-coin projects have so far failed to surface quality content, and proposes a token-linked creator DAO as an alternative. He says the core problem is quality rather than volume, and that Substack has succeeded because its launch was hands-on, seeding the platform with selected high-quality creators and offering some of them revenue guarantees. His proposal is a non-token-based DAO with a few dozen to about 200 hand-picked members who vote others in and out, with creators admitted to it having a portion of their proceeds used to burn their own creator coins. Under this design, speculators would effectively predict which creators the DAO will admit.
- Deep Funding
Deep funding proposes combining two ideas: treating value as a graph, where the question becomes how much of the credit for one project belongs to another, and using distilled human judgement, where an open market of AIs proposes weights for all edges and a human jury spot-checks a random subset. The author says the graph approach has grown as a paradigm because contribution is hard to measure in the abstract, while "local" comparisons such as whether A or B is more valuable to C are more tractable. Under the proposed method, the submissions most compatible with the jury's answers determine the final output. The article links to deepfunding.org for the project.
- vitalik.eth: In 2026, I plan to be fully back to decentralized social
Vitalik Buterin says he plans to be fully back on decentralized social in 2026, arguing that better mass communication tools need more competition, which decentralization enables through a shared data layer that anyone can build clients on. He says he has already used decentralized social this year, reading and posting through Firefly, a multi-client covering X, Lens, Farcaster and Bluesky. He criticizes crypto social projects that treat adding a speculative coin as innovation, citing Substack as proof that money and social can mix well when it means subscribing to creators rather than creating price bubbles. He praises the Aave team for stewarding Lens and says the incoming team is genuinely interested in the social side, and he encourages others to spend more time on Lens, Farcaster and the wider decentralized social world.
- @vitalik.eth: You don't have to agree with me on anything to use Ethereum
Vitalik Buterin argues that people can use Ethereum without agreeing with him on applications, trust assumptions, politics, DeFi, AI, or other personal views. He states that he does not represent the whole Ethereum ecosystem, and that criticizing applications he considers "corposlop" is not censorship, since he is free to criticize but not to shut them down. He contends that principles should shape which applications get built on a protocol and that neutrality belongs to protocols like HTTP, Bitcoin, and Ethereum rather than to people. He adds that a decentralized protocol cannot belong to just one metaverse, and that a metaverse's borders are fuzzy, so people can align with one group on some issues and not others.
- Poor man‘s 2-out-of-3 seed phrase split
- @vitalik: proposal for secret UN voting to reduce global power dynamics
- vitalik.eth: 2023 roadmap vs Strawmap shows new priorities in privacy and PQ
Vitalik Buterin compares his 2023 Ethereum roadmap diagram with the current Strawmap and argues that the newer plan shows changed priorities, especially in privacy and post-quantum security. He says quantum safety has been moved up in order, while VDFs and many EVM improvements were deprioritized, and Verkle trees and state expiry were replaced with newer constructions. The Strawmap adds items absent from 2023, including strong privacy work, leanSPHINCS signatures, native rollups, blob and gas futures, and a more open design space for the EVM. He also says formal verification, made feasible by modern AI tools, and STARKs are now central to the protocol's technical direction.
- vitalik.eth: eth.limo DNS registrar compromised, avoid until resolved
- vitalik.eth: donated 128 ETH each to Session and Simplex for metadata privacy
Vitalik Buterin says he donated 128 ETH each to Session and Simplex, two messaging apps he describes as pushing permissionless account creation and metadata privacy forward. He urges readers to download and use both apps, while noting that neither is yet optimal in user experience or security. The post says strong metadata privacy requires decentralization, which is hard, and that multi-device support and Sybil or DoS resistance add further difficulty.
- @vitalik.eth: how Railgun's privacy pools filter criminal deposits
Vitalik Buterin highlights that Railgun's privacy pools mechanism filters out deposits linked to criminal activity without using snooping or backdoors. According to the post, anyone can deposit into Railgun, and algorithms then spend a 1 hour period checking whether the deposit likely came from activity they consider criminal. Deposits that pass can be withdrawn privately using ZKPs after 1 hour, though waiting longer is advised for a better anonymity set. Deposits that fail can only be withdrawn to the depositor's own address, with no risk of freezing or seizure, but without access to the privacy pool.
- vitalik.eth: I withdrew 16,384 ETH to fund open-source verifiable tech stack
Vitalik Buterin says he has withdrawn 16,384 ETH to fund an open-source, secure and verifiable full stack of software and hardware that protects personal and public life. He says the funds will be deployed toward these goals over the next few years and that he is exploring secure decentralized staking options to direct more staking rewards toward them over the long term. The post places this alongside the Ethereum Foundation's period of mild austerity, which it describes as a way to meet its roadmap goals and sustain its mission. It also names as examples the Vensa announcement on open silicon, the uCritter project, air quality work, and donations to encrypted messaging apps.
- Rethinking Ethereum's L2s: Beyond Branded Shards in a Scaling L1 Era
Vitalik Buterin argues that L2s should stop being treated as literal "branded shards" of Ethereum, since L1 is now scaling directly and L2 progress to stage 2 has been slower than expected. He describes a spectrum of L2 designs, from chains backed by Ethereum's full faith and credit to options with weaker connections, and suggests each L2 should offer a distinct value beyond scaling, such as non-EVM privacy VMs or ultra-low-latency sequencing. L2s handling ETH or Ethereum-issued assets should be at least stage 1. He also proposes a native rollup precompile that verifies ZK-EVM proofs, which would auto-upgrade with Ethereum and make trustless interoperability easier.
- @vitalik: Two core missions for ethereum's future growth
- vitalik.eth: testing mobile offline local knowledge apps
Vitalik.eth tested several mobile offline local knowledge apps, noting they have improved over a self-built attempt from two months earlier. He said they remain slower and less effective on difficult questions than models that run on a laptop. The apps performed weakest on specialized travel queries, such as asking for the best vegan restaurants in the city he was in, where none did well. He expressed hope that offline apps will soon let users look up facts without internet access.
- vitalik.eth: Surveillance enables small-coalition regimes to survive forever
Vitalik Buterin argues that surveillance is bad not because it is "dystopian" but because it shifts the balance of power between individuals and the state, letting a small coalition of security forces hold power indefinitely. He cites The Dictator's Handbook's distinction between large and small coalition governments and warns that dictatorship combined with surveillance and automated police could let a regime survive with a coalition of one. He also argues that Western surveillance differs in kind: it exercises medium control over a large area, enabling global projection of power, such as the US pressure on EU and ICC officials he mentions. He concludes that privacy generally helps whoever is weaker.
- AMA with Vitalik
- "The Purge" (Ethereum code and protocol simplification) is happening!
- @vitalik.eth: anon billboard with moderation demo on Aztec
- @vitalik.eth: EF will focus on CROPS, longevity over breadth, provable security
Vitalik Buterin writes that the Ethereum Foundation (EF) is refocusing on CROPS (censorship resistance, open, private, and secure systems) and longevity over breadth, which means selling less ETH. He says the EF has about 0.16% of all ETH and that its original limited scope was fully completed in 2022, so it should act as one node among others rather than the center of Ethereum. Buterin lists goals including provably bug-free Ethereum, available chain consensus, and intermediary minimization, with work such as FOCIL, EIP-8141, and Kohaku. He says the EF will be a smaller, more opinionated organization, and that he is only one board member and his own power there will decrease.
- @vitalik.eth: Getting increasingly bullish on vibe-coding in Lean
- @vitalik.eth: Ethereum should build sanctuary tech, not remake the world
Vitalik Buterin argues that Ethereum should build "sanctuary technologies" rather than try to remake the world. These are free, open-source tools that let people live, work, communicate, manage risk and collaborate in ways that stay robust to outside pressures. He frames the goal as de-totalization: creating digital islands of stability and interdependence that cannot be weaponized, rather than total control by any one winner. He cites Starlink, locally-running open-weights LLMs, Signal and Community Notes as examples of liberating technologies, and says Ethereum's role is to provide "digital space" for cooperation, such as money and multisigs that outlive any single key.
- @vitalik.eth: updated view on ethereum + AI intersections and d/acc future
Vitalik Buterin argues that Ethereum and AI should be integrated around a d/acc direction, meaning deliberately chosen positive outcomes rather than undifferentiated acceleration, and that his updated focus is on trustless and private interaction with AI. He lists local LLM tooling, ZK-payments for API calls so users can avoid linking their identity across calls, and client-side verification of cryptographic proofs and TEE attestations as parts of that work. He also describes Ethereum as an economic layer for AI, covering API calls, bots hiring bots, security deposits, and ERC-8004 reputation ideas. He says LLMs could remove the limits of human attention that have constrained prediction markets, quadratic voting and other governance mechanisms.
- vitalik.eth: We need more DAOs - but different and better DAOs
Vitalik Buterin argues that more DAOs are needed, but designs different from the common token-holder voting treasury, which he calls inefficient, vulnerable to capture, and unable to mitigate the weaknesses of human politics. He names several needs DAOs could meet: better oracles for stablecoins and prediction markets, onchain dispute resolution, maintaining lists such as known scams and token addresses, funding projects quickly, and long-term project maintenance. He proposes ZK for privacy, AI to address decision fatigue, and consensus-finding communication tools, and says projects should treat oracle and communication-layer design as 50% of their work.
- @vitalik.eth: Prediction markets need interest payments to unlock hedging
Vitalik Buterin argues that prediction markets need to pay interest before hedging use cases can open up. He says major prediction markets currently mostly don't pay interest, which makes them unappealing for hedging because participating means giving up a guaranteed 4% APY on the dollar. He expects hedging use cases and trading volumes to grow once this is solved.
- Philosophy of the app layer
The author argues that the growth of the app layer makes good social philosophy especially important, because apps are largely special-purpose and their design depends on what people think Ethereum and its apps are for. The argument contrasts C++, a general-purpose language that would likely be no worse if built by someone with bad social views, with Ethereum L1, which is perhaps 50% general-purpose: a person who doesn't believe in decentralization would not add light clients, FOCIL, or account abstraction, though the EVM opcodes might be much the same. Apps are described as roughly 80% special purpose, so having good ideas about their purpose becomes crucially important.
- vitalik on beam chain
Vitalik Buterin argues that the Beam Chain roadmap, which drew criticism for its "2029" date, is a separate workstream from the near-term work on scaling blobs, and that Ethereum needs a clean, simple protocol that removes large amounts of technical bloat. He says that in 2025 the network needs blobs and L1 cost reductions such as EIP-4444, and in 2026 it needs more blobs. He names Justin Drake, Dankrad Feist and Francesco (fradamt) as contributors who are pushing this research forward. By 2030, he says, Ethereum will also need long-term robust economics and staking decentralization.
- Vitalik‘s last cast was 25d ago
- vitalik.eth: onchain mechanism design needs two layers
Vitalik Buterin argues that onchain mechanism design is mostly going to follow a two-layer pattern: a prediction-market-style execution layer and a separate preference-setting layer. He says the market layer is open and accountable, since anyone can trade and good or bad decisions earn or lose money. The preference-setting layer should be decentralized and pluralistic, and he argues it cannot be token-based because token owners are not pluralistic and anyone can buy 51% of them, so votes there should be anonymous, ideally MACI'd to reduce collusion risk. He also says a replaceable centralized executive can be used at the execution layer when simplicity is wanted.
- Using local models + zkAPI + Tor for private AI
The author describes a self-experiment to get personalized diet and exercise recommendations from frontier AI models using their personal health and travel data without leaking private information. A local model, Qwen 3.8 Flash Next, orchestrates the process and writes the queries to remote models, which the author says helped improve the recommendations. The approach uses three privacy layers: the local model to avoid leaking identity through content or writing style, zkAPI to avoid leaking identity through payments, and Tor to avoid leaking it through networking. The author lists deficiencies, including Tor being poorly suited to request-by-request de-linking and adding high latency, the skill file's query strategies being far from optimal, the local model running at 20-30 TPS, and a tradeoff where more privacy limits how much remote models can help.
- vitalik.eth: Qwen 3.8 flash and llama.cpp make local AI models viable
Vitalik Buterin writes that Qwen 3.8 flash is impressive and that llama.cpp has been rapidly improving at running it, and he says local models are close to handling a large share of tasks. He reports benchmark results from his laptop, which uses a Strix Halo chip, with columns for pre-existing prompt, new prompt, generated tokens, input tokens per second, and output tokens per second. He also suggests that using a local model to orchestrate queries to powerful remote models could become viable, so that personal information does not leak to those models.
- @vitalik.eth: We should revisit beacon/execution client separation
Vitalik Buterin argues that Ethereum should be open to revisiting the separation of beacon and execution clients, because running two daemons that must talk to each other is much harder than running one. He says the goal is good user experience for self-sovereign use of Ethereum, which in many cases means running your own node, and that the current setup adds needless complexity. In the short term he suggests a more standardized wrapper for running any clients in Docker and connecting them easily, and he welcomes the unified node in status-im's nimbus-eth1 pull request 3646. Over the longer term he proposes revisiting the whole architecture once lean consensus is more mature.
- vitalik.eth: rethinking democratic tools for chaotic eras vs stable eras
Vitalik Buterin argues that democratic tools should be built differently in chaotic eras than in stable ones, favoring consensus-finding tools over binding mechanisms. He says the 2000s and 2010s were stable, when ambitious ideas like global UBI and quadratic voting seemed possible, while the 2020s are chaotic, with interventions driven by raw power-grabbing. He proposes tools such as Pol.is-style consensus mapping, anonymous voting, and assurance-contract-style statements that become public once they reach a support threshold, giving distributed groups a voice that powerful actors must consider. He also raises the question of how Iranian people can gain a collective voice with hard power during the ongoing Iran war.
- @vitalik.eth: prediction markets need hedging, not dopamine-driven bets
Vitalik Buterin argues that prediction markets are drifting toward short-term crypto price bets and sports betting, which offer entertainment value but little long-term or societal information value, and that they should instead be pushed toward hedging. He says teams chase these products because they generate large revenue in bear markets. He describes three sources of losing participants: naive traders, info buyers who subsidize markets to learn information, and hedgers who accept negative expected value to reduce risk. He argues that info buying faces a public goods problem, so hedging is the more sustainable path. He proposes replacing fiat currency with personalized baskets of prediction market shares tied to each user's expected expenses, denominated in assets such as interest-bearing fiat, wrapped stocks or ETH.
- @vitalik.eth: adversarial governance mechanism design may be key to AI safety
Vitalik Buterin suggests adversarial governance mechanism design theory may be key to AI safety. He compares a 2020 coordination post with an aiprospects Substack piece on preventing AI collusion, describing a shared setup in which a less-sophisticated principal seeks ideal outcomes from more-sophisticated agents, such as humans or stronger LLMs. A key finding is that better outcomes are possible when collusion among agents is limited, since Nash equilibria are abundant while cooperative-game-theory cores are often empty, and this argument should naturally transfer to AI.
- vitalik.eth: 33% chance SNARKs, FHE, and iO reach 1+ε overhead in the future
Vitalik Buterin estimates a 33% chance that, for average real-world computation, SNARKs, FHE, and iO can each be implemented with 1+ε overhead, meaning the added cost of cryptographizing a computation becomes arbitrarily small relative to the base cost for large enough instances. He gives a 60% chance that all three can be done with single-digit overhead, below 10x measured in total energy cost plus amortized compute. He says a good chance exists of reaching one of these, probably SNARKs with single-digit overhead, by the end of this decade, pointing to existing results for specialized hash functions and some LLM inference.
- vitalik.eth: Transaction actions vs dependencies and EIP-8141's role in scaling
Vitalik Buterin argues that recent work on transaction formats, including EIP-8141, gives a clearer way to separate transaction "actions" from "dependencies," so the two can be optimized separately. An action is an effect a transaction has, such as a call that sends ETH, while a dependency is a fact that must hold for the transaction to be valid, such as a signature, a Merkle proof of a UTXO, or a ZK-SNARK or STARK. Pure dependencies can be processed once at the mempool layer and potentially replaced with a STARK, and the text says Ethereum's scaling strategy requires making statically analyzable features cheaper than dynamic ones. The article states that over 90% of Ethereum's activity by volume does not need dynamic, flexible execution. It describes a general-purpose transaction type as a natural interface layer for these changes and calls EIP-8141 done well preparation for the next few years of scaling.
- @vitalik.eth: Lean Ethereum roadmap update from Berlin researcher meeting
Vitalik Buterin reports on an Ethereum researcher meeting in Berlin that updated the protocol's long-term roadmap, with the updated strawmap published at strawmap.org. He describes "Lean Ethereum" as a set of improvements rolling out over three or four years, which he calls the third major iteration of Ethereum after the Merge. Planned changes include recursive STARK verification, quantum-safe replacements, a new consensus design, multidimensional gas and changes to state. He also gives an illustrative 2030 scenario of 2 TB of present-day-style state and 100 TB of new-style state.
- @vitalik.eth: options-based index tracking assets being built urges verification
Vitalik Buterin says options-based index tracking assets are already being built and urges that any of them be formally verified before reaching mainnet. He points readers to an Ethereum Research thread where people are developing different versions of the idea, which builds index-tracking assets on top of options instead of debt. He also hopes the Vyper team and the Verity project, with Fricoben, can help with verification and notes that now is a good time to think about robustness-optimized oracles.
- @vitalik.eth: Ethereum L1 protocol research taking leaps forward in 2026
- @vitalik.eth: Good incentives compensate costs, bad ones attract users who leave
Vitalik Buterin argues that good crypto incentives compensate for temporary costs of an immature project, while bad ones attract users who leave once the payments stop. He says incentives that bring in users who would not use a mature version of the product are bad, and gives paying people to make tweets as a clear example, since it draws users focused on gaming the system. He also argues that rewarding many users in quantity can fail to grow community quality, which he says matters more in the 2020s outside DeFi. He adds that the bulk of effort should go into building a useful app, with user acquisition done through means other than indiscriminate payments.
- vitalik.eth: stop making copypasta EVM chains, build something new
Vitalik Buterin argues that building yet another EVM chain with an optimistic bridge to Ethereum, or an alternative L1 without one, is a dead end that copies existing designs instead of bringing something new. He suggests new chains should offer distinct features such as privacy, app-specific efficiency, or ultra-low latency, and notes that L1 scaling will supply lots of EVM blockspace, though AI workloads may need more blockspace and lower latency than a scaled L1 can offer. He also argues that a project's public connection to Ethereum should match its actual technical connection, with app chains that are deeply tied to L1 described as Ethereum applications and those that are not described directly as separate projects.
- @vitalik.eth: Protocol simplicity is key to trustlessness and self-sovereignty
Vitalik Buterin argues that protocol simplicity is essential to trustlessness, passing the walkaway test, and self-sovereignty, and that Ethereum needs an explicit simplification process to counter the bloat caused by backwards-compatibility pressure. He says a protocol of hundreds of thousands of lines and layered advanced cryptography fails all three tests, and that complexity also adds security risk. He lists three simplification metrics: minimizing lines of code, avoiding unnecessary complex dependencies such as isogenies, and adding invariants like EIP-6780 and EIP-7825. He also points to piecemeal and large-scale garbage collection, including the Glamsterdam gas cost reforms and Lean consensus, and to demoting little-used features into smart contract code.
- @vitalik.eth: We need better decentralized stablecoins with improved oracles
Vitalik Buterin argues that better decentralized stablecoins need three things: a tracking index better than the USD price, an oracle design that a large pool of money cannot capture, and a solution to staking yield competing with collateral use. He says tracking USD is acceptable short term, but over a 20-year timeline, even moderate hyperinflation could make price independence from USD important. Without decentralized oracles, he writes that cost of capture must exceed protocol token market cap, which implies value extraction above the discount rate and is bad for users. He also lists three paths for the staking problem: lowering staking yield to about 0.2%, creating a higher-yield staking category without the same slashing risk, or making slashable staking usable as collateral.
- vitalik.eth: Ethereum as Linux and BitTorrent, serving purists and billions
Vitalik Buterin compares Ethereum to Linux and BitTorrent, arguing that Ethereum should serve purist users while also reaching mass adoption. He writes that BitTorrent combines decentralization with mass scale, and that Ethereum aims to do the same with consensus. He notes that Linux is free and open source, is depended on by billions of people and enterprises, and is used by governments, while distributions such as Arch serve highly purist, minimalistic users. He argues that Ethereum L1 should serve as a financial home for people seeking autonomy, and that this is compatible with broad value for large numbers of people.
- vitalik.eth: ZKEVMs and PeerDAS shift Ethereum into a new kind of network
Vitalik Buterin argues that ZK-EVMs and PeerDAS together make Ethereum a new kind of decentralized network that combines decentralization, consensus and high bandwidth. He says PeerDAS, a form of data availability sampling, is live on mainnet, while ZK-EVMs are at alpha stage with production-quality performance and remaining work on safety. He expects small portions of the network to use ZK-EVMs in 2026, with ZK-EVMs becoming the primary way to validate blocks in 2027-30. He also describes distributed block building, in-protocol or through distributed builder marketplaces, as a further goal.
- vitalik.eth: the year we take back lost ground in computing self-sovereignty
Vitalik Buterin argues that 2026 should be the year people take back lost ground in computing self-sovereignty, a goal he says applies well beyond blockchains. He describes switching his own software: moving to the open-source encrypted docs app Fileverse and to Signal as his primary messenger in 2025, then in 2026 replacing Google Maps with OpenStreetMap via OrganicMaps, and Gmail with Protonmail. He says local LLM setups still need work to become well-integrated, and that heavy-usage tasks could combine per-query zk payments, TEEs, and local filtering of sensitive details, with efficient FHE as the longer-term goal.
- @vitalik.eth: America's True Mission vs. Today's Racist Immigration Lies
- vitalik.eth: Keyed nonces as Ethereum's first specialized state scaling strategy
Vitalik Buterin argues that keyed nonces could be Ethereum's first specialized state scaling strategy, creating storage types optimized for specific use cases and restricted so they remain usable at extreme scale. Using in-protocol nullifiers as the example, he calculates that 2000 transactions per second of privacy-preserving activity for eight years would produce about 500 billion nullifiers, which cannot be pruned. He states that a dedicated nullifier store is easier to keep decentralized than growth in the current state, because nullifiers can be sharded across nodes and checked with Bloom filters that reduce the requirement to about 8 bits per nullifier. Without these techniques, builders would need to download the full 16 TB, so the article proposes moving most usage into specialized state while keeping fully dynamic state for applications such as DeFi.
- vitalik.eth: Ethereum should rethink apps from first principles, not iterate
Vitalik Buterin argues that Ethereum should rethink its application layer from first principles rather than iterate on existing designs, while keeping core properties of censorship resistance, open source, privacy and security (CROPS) fixed. He cites privacy as a first-class consideration, which he says implies a radically different application stack, and points to growing work on network-level privacy. He also suggests rethinking the role of L2s and asks builders to imagine the chain has zero usage and to build the first apps anew.
Most discussed
- vitalik.eth: How I would do creator coins
- @vitalik: proposal for secret UN voting to reduce global power dynamics
- vitalik.eth: the year we take back lost ground in computing self-sovereignty
- vitalik.eth: 2023 roadmap vs Strawmap shows new priorities in privacy and PQ
- vitalik.eth: eth.limo DNS registrar compromised, avoid until resolved