20 Matching Annotations
  1. Last 7 days
    1. These system prompt updates do not apply to the Claude API.

      这里有一个关键的非共识观点:Anthropic刻意保持API和界面行为的不一致性。虽然表面上API提供了更多控制权,但这种分裂意味着API用户可能会错过重要的行为改进和安全更新,这实际上创造了两个不同的'Claude'生态系统。

  2. Apr 2026
    1. You can see a list of supported browsers in our Help Center.

      这个看似常规的提示实际上揭示了Web生态系统的碎片化问题。平台需要明确列出支持的浏览器,暗示了不同浏览器实现标准的差异,以及开发者需要为不同环境适配的额外负担。这种碎片化是Web开发持续面临的挑战。

    1. Most skills require you to install a dedicated CLI. But what if you aren't in a local terminal? ChatGPT can't run CLIs. Neither can Perplexity or the standard web version of Claude.

      这个观察揭示了Skills模式的一个致命弱点:环境局限性。作者指出了一个令人惊讶的事实:许多流行的AI平台实际上无法运行CLI工具,这使得依赖CLI的Skills在这些环境中完全失效。这不仅是技术限制,更是生态系统的重大分裂。

    1. Each platform surfaces different vulnerabilities, making it difficult to establish a single, reliable source of truth for what is actually secure.

      这一观察揭示了AI安全工具的碎片化问题,不同AI平台发现的漏洞各不相同,导致难以确定真正的安全状态。这种不确定性不仅增加了防御难度,还可能引发安全评估的混乱,需要建立新的行业标准来应对AI时代的安全挑战。

    1. DeepSeek is the only product that bridges the divide.

      DeepSeek 同时在中国、俄罗斯、美国获得显著用户,在技术分化的世界中极为罕见。它不仅是产品,更是地缘政治缝隙中的独特存在——既规避西方制裁,又突破中国的封闭性。这种「跨界」属性是护城河也是风险源:当三个监管体系冲突时,它能否维持这种微妙平衡?

  3. Nov 2025
  4. Oct 2024
  5. May 2024
    1. DNA fragmentation by restriction digestion prior to dropletgeneration enables optimal accuracy by separating tandemgene copies, reducing sample viscosity, and improving templateaccessibility for input samples >66 ng per well

      NEB says

      Digestion is recommended whenever DNA input is greater than 75 ng Source

      NEB says that biorad recommends these enzymes: AluI, CviQI, HaeIII, HindIII-HF, MseI

      More guidelines

      • Assemble ddPCR reactions at room temperature, this allows the restriction enzyme to digest the gDNA during the reaction set-up period
      • Prepare reaction mixes as you would for a standard ddPCR reaction. Add 0.5–1 μL of each restriction enzyme (5–20 units, depending on enzyme concentration) to the reaction mixture
      • After set-up, simply continue droplet generation as normal
      • Restriction enzyme will be inactivated during first PCR denaturation step
  6. Dec 2023
    1. what i would call a kind of epistemic fragmentation where where we're losing the ability as societies to have agreement on very basic facts and to the extent that we don't agree on 00:36:41 basic facts about the nature of the world and the nature of the challenges we face it's very hard to to solve those problems effectively democracy for instance can't effective effectively function 00:36:52 if if the people who are talking to each other don't actually agree on what the problems are
      • for: definition epistemic fragmentation, no agreement on basic facts, political polarization, adjacency - epistemic fragmentation - polarization - democracy

      • definition: epistemic fragmentation

        • when major parts of society disagree on basic facts
      • adjacency between
        • epistemic fragmentation
        • polarization
        • democracy
      • adjacency statement
        • democracy is critical to solving our challenges but it won't function if there is so much epistemic fragmentation that we can't agree on basic facts
  7. Aug 2023
    1. We need mass innovation in design of social tools that help us bridge fragmentation and polarization, bring diversity into our media landscapes and help find common ground between disparate groups. With these as conscious design goals, technology could be a powerful positive force for civic change. If we don’t take this challenge seriously and assume that we’re stuck with mass-market tools, we won’t see positive civic outcomes from technological tools.”
      • for: quote, quote - Ethan Zuckerman, quote - fragmentation and polarization, Indyweb - support, MIT Center for Civic Media, Global Voices
      • quote
        • We need mass innovation in design of social tools that help us
          • bridge fragmentation and polarization,
          • bring diversity into our media landscapes and
          • help find common ground between disparate groups.
        • With these as conscious design goals,
          • technology could be a powerful positive force for civic change.
        • If we don’t take this challenge seriously and assume that we’re stuck with mass-market tools,
          • we won’t see positive civic outcomes from technological tools.”
      • author
        • Ethan Zuckerman
          • director of MIT’s Center for Civic Media and
          • co-founder of Global Voices
  8. Jun 2023
  9. Feb 2023
  10. Dec 2022
    1. I don't know how this will look like. What I do think is it will come to cultural identity. What is the cultural identity? And that's what we will all gravitate to, and we'll gravitate.

      !- future global fragmentation : by culture - Michaux believes people will fragment in the future along cultural boundaries as we move through tumultuous transition. This makes sense as ingroups will naturally form - this should be further explored to explore implications: - will we get political polarization? At what level? National, regional, city / community scale? - what implications will this have on cooperation and sharing? will it create policy gridlock? Will it become even more urgent to educate everyone on a Deep Humanity type of open praxis that finds common human denominators (CHD)?

  11. Aug 2022
  12. Jul 2021
    1. For much of Americanhistory, people were educated in a wide range of (often highlyeccentric) settings. !is was generally perceived as a problem, andefforts at standardization kicked in, reaching their peak in the1960s. Since then we have seen increasing fragmentation, withordinary public schools, charter schools, magnet schools, variouskinds of private schools, homeschooling, unschooling—but all ofthese work on the same platforms;

      Interesting to note the time period of this peak, which broadly coincides with Brown vs. Board of Education and desegregation in the 1970s. Did the fluorescence of these others begin as a means of better segregating students either racially or economically?

  13. Oct 2020
    1. Think of this essay as a series of strongly held hypotheses; without access to the types of data which i’m not even sure exists, it’s difficult to be definitive. As ever, my wise readers will add or push back as they always do.

      Push back, sure, but where? Where would we find this push back? The comments section only has a few tidbits. Perhaps the rest is on Twitter, Facebook, or some other social silo where the conversation is fraught-fully fragmented. Your own social capital is thus spread out and not easily compiled or compounded. As a result I wonder who may or may not have read this piece...

  14. Jul 2020
  15. May 2020