888 X 動態摘要｜08/12
生成時間: 2026-08-12 05:02:00
總結
Vitalik Buterin闡述了以太坊未來以量子安全、隱私優先、AI輔助安全驗證及底層架構革新為核心的宏大路線圖，同時醫療AI在視訊診斷方面展現突破性進展。
今日重點
1. 以太坊量子安全戰略
- 人物: Vitalik Buterin (以太坊創始人)
- 時間: Aug 10
- 熱度: 👀 482,727
- 觀察: 以太坊將量子安全優先級提高，計劃引入leanSPHINCS等技術，積極應對後量子加密挑戰。
- 意義: 確保區塊鏈基礎設施的長期安全性與韌性，對於吸引大型機構應用和維護生態系統信任至關重要。
2. 協議級隱私強化
- 人物: Vitalik Buterin (以太坊創始人)
- 時間: Aug 10
- 熱度: 👀 482,727
- 觀察: 以太坊正將強隱私提升為核心開發重點，導入如鍵控Nonce、FOCIL及精簡隱私池等具體機制。
- 意義: 滿足用戶對匿名性和數據保護的日益增長需求，有望擴大以太坊在金融和企業應用場景中的採用。
3. AI助推安全驗證
- 人物: Vitalik Buterin (以太坊創始人)
- 時間: Aug 10
- 熱度: 👀 482,727
- 觀察: 透過AI工具實現全面的形式化驗證，以太坊的協議規範將更精簡，大幅提升安全設計的嚴謹性。
- 意義: AI技術被確立為確保區塊鏈核心協議安全與可靠性的基石，降低潛在漏洞風險，增強平台穩健性。
4. 新型金融衍生品構想
- 人物: Vitalik Buterin (以太坊創始人)
- 時間: Aug 10
- 熱度: 👀 482,727
- 觀察: 以太坊提出Blob和Gas期貨的新概念，旨在為未來的協議引入費用預測和對沖機制。
- 意義: 這些新的鏈上金融原語可能為開發者和用戶提供更穩定的交易成本管理，並催生新的市場服務。
5. EVM底層架構革新
- 人物: Vitalik Buterin (以太坊創始人)
- 時間: Aug 10
- 熱度: 👀 482,727
- 觀察: 以太坊考慮將EVM轉化為IR，並引入leanISA或RISC-V等更高效的指令集，以優化協議效率和開放性。
- 意義: 這項潛在的底層技術變革將大幅提升以太坊的性能與可擴展性，對現有EVM生態及新技術採用產生深遠影響。
6. 醫療AI視訊診斷
- 人物: Eric Topol (頂尖心臟病專家)
- 時間: 3h
- 熱度: 👀 21,734
- 觀察: 醫學AI研究顯示，結合視訊功能的AI診斷系統在遠程醫療諮詢中，診斷準確性優於純文本或人工判斷。
- 意義: 證明AI在遠程醫療領域的巨大商業化潛力，尤其在提升診斷效率和準確性方面，將驅動醫療科技投資。
原始動態
重點 | Vitalik Buterin
- 時間: Aug 10
- 熱度: 👀 482,727
- 原文: I updated my 2023 roadmap diagram to overlay where the items that were there sit in the current Strawmap ( "). In general, a lot of overlap, but: * Some things got reshuffled in order (eg. quantum safety up-prioritized) * Some things deprioritized (eg. VDFs; many EVM improvements) * Some things replaced with superior constructions (eg. Verkle -> unified BT -> PBT; state expiry -> new state types) What's most striking, however, is that some completely new things are in the strawmap that are NOT in this diagram, because they were not in the 2023 roadmap at all. These reflect changing priorities. Notably: * First-class attention to strong privacy. This covers: keyed nonces and recent roots, aspects of FOCIL, lean privacy pool & wormholes * Aggressive scaling in the context of post-quantum. This covers: leanSPHINCS signatures and aggregation, zkzk frames (see" ") * Lean-ification of the spec, to assist in formal verification (full FV of everything is only possible because of modern AI) * Blob and gas futures (this idea just didn't exist back in 2023) * Native rollups (SNARKs were nowhere near mature enough to even consider this back in 2023) * A more open design space for the \"future of the EVM\". zkzk frames already implies that the protocol will expose to users some ISA that's not the EVM - current leading candidates are leanISA and RISC-V. These ISAs are more simple, modern and efficient than the EVM. Once they're there, why not expose them to developers everywhere? (And then, why not turn the EVM into being an IR on top of that ISA, instead of an enshrined feature massively complicating the base protocol?) Though much of the deeper exploration here is too early even for the strawmap. * New state types are not just a replacement for expiry, they're a fundamentally different paradigm to how Ethereum does scaling A common theme in scaling, found in both state types and zkzk frames (both new ideas), is that instead of trying to maximally scale ALL ethereum activity, we try to create specialized mechanisms that have more restrictive properties that make them more scaling-friendly, while supporting the heaviest loads incurred by users and applications today (eg. token transfers, swaps) and tomorrow (eg. privacy protocols). The other common theme is treating STARKs and AI-accelerated FV as first-class objects, that we are okay betting the technical future of Ethereum on. There are recursive STARKs in many layers of the protocol, one particular primitive (the \"aggregate to union verified dependencies\" primitive) is expected to be used in three places in the protocol: EL, CL and DL. This can only be safe with formal verification, which is itself only feasible with modern AI tools. In general, many steps forward in maturity. And a huge amount of hard work by many dozens of Ethereum researchers and developers on all of these features. Ethereum will be quantum-safe. Ethereum will put users' privacy first. Ethereum will be secure. Ethereum will be censorship-resistant. Ethereum will be highly performant and scalable while satisfying the above. And Ethereum will be Lean."
其他 | Elon Musk
- 時間: 4h
- 熱度: 👀 477,047
- 原文: The word "steampunk" was coined in April 1987 by American science fiction writer K. W. Jeter. He proposed it jokingly in a letter to Locus magazine to describe the alternate-history novels set in the 19th century that he was writing alongside his fellow authors Tim Powers and James Blaylock.
其他 | Nassim Taleb
- 時間: 6h
- 熱度: 👀 162,723
- 原文: We can approach it above by discussing multivariate probability distributions. As to random variables, say x and f(x) where f(x) is not an affine transformation, they will decorrelate while being deterministically dependent. This is the problem with IQ "science" retards who know no math.
其他 | Elon Musk
- 時間: 1h
- 熱度: 👀 889,210
- 原文: Moonbase Alpha will be so cool. And we want to make it so that anyone can go to the Moon if they want.
其他 | Elon Musk
- 時間: 1h
- 熱度: 👀 188,704
- 原文: RT ": I’m picking commercial flights based on starlink, not total time, route or price This new low cost Japan carrier" …
其他 | Ray Dalio
- 時間: 22h
- 熱度: 👀 73,279
- 原文: Many of you have asked about how I've learned to deal with setbacks, so I’m sharing a list of principles that have helped me over the years.
其他 | Ray Dalio
- 時間: 6h
- 熱度: 👀 72,281
- 原文: People keep cash because they think it's safer, but it's not. You're actually losing money because of inflation. I explain my thoughts on why your money could be devaluing in my recent conversation with . youtube.com
其他 | Nassim Taleb
- 時間: 22h
- 熱度: 👀 1,147,631
- 原文: I don't think I've laughed harder, to the point of feeling a stich on my side.
其他 | Elon Musk
- 時間: Aug 10
- 熱度: 👀 1,693,287
- 原文: Toch benieuwd wat "van deze kwalificatie zou vinden (U mag raden wie 'm maakt en waar): \"Premier Jetten moet Yesilgöz, en alle andere bewindslieden van zijn minderheidskabinet, verbieden om zich op sociale media uit te laten over alles dat buiten de eigen portefeuille valt. Daarnaast moet men natuurlijk sowieso geen mededelingen meer doen op X, een platform dat wordt gedomineerd door desinformatie, en dat in handen is van de rijkste man ter wereld, die bovendien een totale fascist is.\"
其他 | Jeff Dean
- 時間: 20h
- 熱度: 👀 260,494
- 原文: Legendary! Thanks for sharing the behind-the-scenes "!"
其他 | Huanusa
- 時間: 21h
- 熱度: 👀 145,780
- 原文: 这位中国高中生在教室里勇敢的说出：不喜欢共产党，也不信共产主义！ 随后被学校开除学籍... 估计这位勇敢少年如果不离开中国话 这一生恐怕都会被体制骚扰折磨 在中国如果你有正义感，有信仰，善良-勇敢-诚实 大家可能都知道后果...这就是粉红眼中的盛世！！
重點 | Eric Topol
- 時間: 3h
- 熱度: 👀 21,734
- 原文: "A new medical AI study demonstrates the diagnostic accuracy advantage of using video (AMIE) over pure text for patient telehealth consultations. A randomized trial of 100 clinical cases, 30 physicians (PCPs), w/ patient actors. 3 arms: AI with video, PCPs with video, text only; main results in Figure (preprint)"
其他 | Vitalik Buterin
- 時間: 2h
- 熱度: 👀 86,421
- 原文: Impressive work! For comparison, an H100 can do roughly 100-200 tok/s of Muse 30B for raw inference single-thread, going up to low thousands of tok/s with a large number of threads - and I am sure that for the massively-multi-threaded case they can optimize the prover further. So we roughly, sort of, have single-digit (<10x) overhead for LLM proving! Next step is getting single-digit overheads for FHE, and then ultimately vFHE (aka STARK * FHE). A crazy ambitious milestone given present FHE overheads, but because of how highly structured and almost-linear LLM inference is, it's closer to the realm of possibility than you might think. Single-digit-overhead all the things. firefly.social
其他 | Huanusa
- 時間: 16h
- 熱度: 👀 22,981
- 原文: 2013年， 100美元⚡️ 讲座来了4个人，走了1个。 留下的3个，现在财富自由 感慨有些东西和机遇都是埋在无人问津时 不只是金融，独立思考是人生最重要的一堂课 大部分人都错过了
其他 | Ray Dalio
- 時間: Aug 10
- 熱度: 👀 24,071
- 原文: I had a blast being out until 4am with my family and friends at UNVRS in Ibiza for my 77th birthday and thought I’d pass it along to turn you onto it, in case you’re in the neighborhood.
其他 | Yann LeCun
- 時間: Aug 10
- 熱度: 👀 4,233,738
- 原文: Today we're also opening the weights for Muse Glimmer, a great 30B parameter dense model that can run locally. Soon we'll also release the weights for Muse Spark 1.2, our latest foundation model. Meta is a strong supporter of open source and I'm proud of these releases. Congrats to and the MSL team for all your great work on these models.
其他 | Yann LeCun
- 時間: Aug 10
- 熱度: 👀 3,686,580
- 原文: I believe everyone should have access to superintelligence, and I wrote a long piece about Meta's philosophy and values for building a positive future for everyone.
其他 | Peter Steinberger
- 時間: Aug 10
- 熱度: 👀 223,072
- 原文: I don't think many humans are going to be reading or writing code in 5 years, so I don't really think it matters.
其他 | Peter Steinberger
- 時間: 18h
- 熱度: 👀 195,293
- 原文: Funny how that headline is about OpenClaw and not Claude. As if the harness could meaningfully prevent a determined user.
其他 | Nassim Taleb
- 時間: Aug 10
- 熱度: 👀 205,117
- 原文: Uncorrelated random variables are not necessarily independent.
其他 | Eric Topol
- 時間: 2h
- 熱度: 👀 12,243
- 原文: "First time this has been assessed: Can AI do a video consult with a patient better than a primary care physician?"
其他 | Sean Kelly
- 時間: 23h
- 熱度: 👀 10,570,620
- 原文: "🚨 JUST IN: Claude models will now have invisible watermarks embedded in ALL text, and ALL metadata attached to files…"
其他 | Huanusa
- 時間: 8h
- 熱度: 👀 28,432
- 原文: 三年大饥荒时期...
其他 | Eric Topol
- 時間: 5h
- 熱度: 👀 4,550
- 原文: Links to our recent view on biological clocks, highlighting organ and cell proteomics to track aging, noting the immune system and brain prominent impact on health outcomes
其他 | Huanusa
- 時間: Aug 10
- 熱度: 👀 5,756
- 原文: 沃伦·巴菲特： “如果没有那份高度专注，我想我在商业上不可能取得如今的成就。
其他 | Nassim Taleb
- 時間: 2h
- 熱度: 👀 22,457
- 原文: The Book is a Book, not Disembodied Information
其他 | Jeff Dean
- 時間: 1h
- 熱度: 👀 3,291
- 原文: And the final couple
其他 | Huanusa
- 時間: 6h
- 熱度: 👀 2,112
- 原文: 轉：没有共产党才有新中国，打倒共产党，习近平下台
其他 | Peter Steinberger
- 時間: 22h
- 熱度: 👀 383,573
- 原文: Must be load-bearing.
其他 | Pierre Levy
- 時間: 20h
- 熱度: 👀 21,608
- 原文: If another model is used to rewrite text generated by new versions of Claude, it will instantly crash and destroy the computer it runs on. Anthropic is currently trying to mitigate that people who read such text sometimes suffer (mild, rare) brain damage, for unknown reasons
其他 | Sean Kelly
- 時間: 5h
- 熱度: 👀 13,124
- 原文: According to a new paper, we might have completely misunderstood the Dunning Kruger effect. Confident people actually do perform better. youtube.com
其他 | Sean Kelly
- 時間: Aug 10
- 熱度: 👀 9,032,582
- 原文: "We asked an unreleased research version of Claude to take a stab at the Riemann hypothesis. It didn’t solve it, but it did make strides on a related problem: it increased the lower bound for the fraction of zeros of the Riemann zeta function that satisfy the hypothesis from 41.6% to 67.2%."
其他 | Eric Topol
- 時間: 3h
- 熱度: 👀 4,006
- 原文: Of course this will require assessment in real world medicine, but it highlights the added dimension of video, capturing parts of the physical examination, and the potential advantage/utility of AI for telehealth consults