放弃大科技:以太坊创始人但erin呼吁2026年是夺回自主计算权的一年。
Abandon Big Tech: Ethereum Founder Buterin Calls 2026 The Year To Reclaim Self-Sovereign Computing

原始链接: https://www.zerohedge.com/technology/abandon-big-tech-ethereum-founder-buterin-calls-2026-year-reclaim-self-sovereign

## 维塔利克·布特林倡导数字自我主权 以太坊联合创始人维塔利克·布特林正在积极转向更注重隐私的数字生活,目标是到2026年“夺回在计算自我主权方面失去的阵地”。他已经做出了重大改变,基本上放弃了中心化平台,转而使用开源和去中心化替代方案。 2025年,他开始使用Fileverse存储文档(一种注重隐私的Google Docs替代品),并使用Signal进行消息传递,理由是其强大的端到端加密。他的2026年计划包括使用OpenStreetMap/OrganicMaps代替Google Maps,以及使用Proton Mail代替Gmail,同时优先考虑去中心化社交媒体。 布特林还在探索本地运行AI模型,他认为随着硬件能力的提高,将数据发送到第三方服务是不必要的。这与隐私倡导者如Naomi Brockwell的观点一致,后者强调本地AI是最私人的选择。 此举反映了人们对数据隐私和政府/企业监控日益增长的担忧,尤其是在欧盟“聊天控制”等提案出台后,并强调了夺回个人数据控制权的重要性。

相关文章

原文

Authored by Christina Comben via CoinTelegraph.com,

Ethereum cofounder Vitalik Buterin declared 2026 to be the “year we take back lost ground in computing self-sovereignty,” starting with his own devices. 

In a Friday post on X, he laid out the software changes he has made to reduce reliance on data-hungry, centralized platforms.

The “two major changes” to the software he used in 2025 were switching “almost fully” to Fileverse, an open-source, decentralized document platform — a kind of privacy-preserving Google Docs — and switching “decisively” to Signal as his primary messaging app.

Signal uses end-to-end encryption by default for all one-to-one and group chats, and stores minimal metadata, meaning only limited information, such as when an account was created or the last date it connected to the service.

Telegram, in contrast, only offers end-to-end encryption in optional “secret chats” and otherwise keeps messages and metadata on its own servers, a model that has drawn scrutiny as law enforcement data requests have increased in countries like France.

Becoming more self-sovereign. Source: Vitalik Buterin

Local AI and self-hosted tools

In 2026, Buterin has moved from Google Maps to OpenStreetMap via OrganicMaps and from Gmail to Proton Mail, while prioritizing decentralized social media.

Buterin also discussed his experiments with locally hosting large language models, arguing that sending all data to third-party services is “unnecessary” when users can increasingly run artificial intelligence tools on their own hardware. 

He said better user interfaces, integrations and efficiency are still needed to make local models a seamless default, but added that there has already been “huge progress” compared with a year ago. 

Privacy advocates see broader shift

His post echoes points made by privacy advocate and NBTV founder Naomi Brockwell, who described running models locally as the most private way to use AI without sending prompts or documents to external servers.

How to use AI privately. Source: Naomi Brockwell

Brockwell has spent years teaching privacy-enhancing behavior to mainstream audiences, arguing that privacy is about autonomy rather than secrecy and encouraging the use of tools like Bitcoin, encrypted messengers and self-hosted services to reduce government and corporate surveillance power.

Buterin’s post also comes amid renewed debate over how much access governments and platforms should have to users’ private communications and metadata.

The European Union’s controversial Chat Control proposal, for example, originally included pre‑encryption scanning of messages to detect abusive material, and prompted warnings from civil liberties groups and technologists that client‑side scanning could undermine trust in encrypted apps.

Progressively swapping out everyday apps for encrypted, open-source and local alternatives is, according to Buterin and other privacy advocates, one way for users to start reclaiming control over their data flows.

Loading recommendations...

联系我们 contact @ memedata.com