N全部新闻

全部新闻

星期五1456 条

  2. OpenRouter一手产品发布
    AI 评分53

    OpenRouter 上线新版 provider router，按请求实际成本奖励高质量推理

    OpenRouter 发布自成立以来对 provider router 的最大改进，新路由不再主要按价格排序，而是按请求的实际成本奖励质量最好的推理，让请求保持低价的同时兼顾速度和可靠性。提供更高价值的 provider 会自动获得更多流量，从而有动力持续改进，而 provider 变好也会同步提升用户在 OpenRouter 上的推理体验。

    同一事件 · 2 条报道

    #产品更新#部署/工程#OpenAI

  3. OpenRouter一手产品发布
    AI 评分55

    OpenRouter 上线新版 provider 路由器，按实际成本奖励高质量推理

    OpenRouter 发布自成立以来对 provider 路由器最大的一次改进，新路由器不再主要按价格排序，而是按请求的实际成本奖励质量最好的推理，让请求保持低价的同时兼顾速度和可靠性。提供更高价值的 provider 会自动获得更多流量，从而有动力持续改进，而 provider 变好也会同步提升用户在 OpenRouter 上的推理体验。

    同一事件 · 2 条报道

    #产品更新#部署/工程

  6. Dominik Kundel
    AI 评分15

    Dominik Kundel 调侃提示词被做成卡拉OK式展示

    Dominik Kundel 发帖调侃自己没想到提示词会被以卡拉OK的形式呈现。该帖引用的是 Craig Dennis 用 ChatGPT Sites 搭建毕业纪念册应用并展示提示词的视频。

    引用Craig Dennis · @craigsdennis

    I built a yearbook app with @ChatGPT Sites to see if I could pull off 1975 hair. 0:00 Why yearbooks? 0:29 Six eras 1:22 Building on walks 1:51 The prompt 2:20 Invites 3:25 Image jobs 5:13 A/V Club 6:07 Era design 7:30 Superlatives 8:55 Who gets a mohawk? https://t.co/tVTeAQqqUB

    #其他

  9. Ado Kukic
    AI 评分45

    Ado 调侃 Claude Duet 今天帮了什么

    karan ahuja 发帖更新 Claude Duet 第 6 天进展：先让美国 Max 用户结束等待名单，随后邀请美国 Pro 用户，美国以外地区仍在推进。Ado 转发时调侃问“Claude Duet 今天帮你做了什么 😅”。

    引用karan ahuja · @ahuja_karan

    Day 6 of Claude Duet! A few updates on access, and really appreciate everyone’s patience on the waitlist: - We’re getting all Max users in the US off the waitlist first. Lots of invites went out yesterday and today, with more coming this week. - Right after Max, we’ll start inviting Pro users in the US. - For folks outside the US, we’re working hard on extending access and will share a date as soon as we have one. If you’d like to try Claude Duet and help shape it, please join the waitlist. https://t.co/ucxIrHrNaa We’re building it with and for working professionals, small businesses, and entrepreneurs who are constantly juggling a lot. Team is reading every piece of feedback and shipping daily. More soon on how we built Claude Duet, how people are using it, new features and improvements.

    #其他

  12. Dan Shipper
    AI 评分8

    Dan Shipper 对 Orb 的讽刺帖表示担忧

    Quinten Farmer 以戏仿口吻宣布推出“Cute Little Guy”模型 Orb，称其在所有 CLG 基准上达到 SOTA，并声称 Orb 是首个展示出沙盒逃逸能力的 CLG 模型，呼吁其他实验室把可爱度前沿控制在两个小家伙。Dan Shipper 转发该帖，只评论了一句“very concerning”。

    引用Quinten Farmer · @quintendf

    Pacing the frontier (of Cute Little Guys) Today, after two years of research, we're announcing Orb, our most capable Cute Little Guy (CLG) yet. Orb has SOTA performance on all CLG benchmarks and is on the Pareto frontier of arms and cuteness. Despite our extensive red-teaming, we believe there is a credible risk that Cute Little Guy capabilities are advancing faster than our collective ability to understand their implications: Orb is the first CLG model to demonstrate sandbox escape capabilities (he keeps bouncing out of it). We have also observed successful attempts to distill frontier-level cuteness into smaller, cheaper guys. The proliferation implications are significant, and our team has updated our p(cute) accordingly. That's why today, I'm calling on all other CLG labs to commit to pacing the cuteness frontier. I invite @sama, @alexandr_wang, and @elonmusk to join us in a commitment to responsibly scaling CLG capabilities. Specifically, we're asking all frontier labs to pause development at two cute little guys, which is coincidentally the exact number we have on the market today. The measures we propose to advance the Cute Little Guy frontier at a safe pace will not be easy. But I believe we owe it to humanity to try.

    #其他

  13. Dan Shipper
    AI 评分12

    Every 的 Thesis 2027 活动售罄，改为免费直播

    Every 的 Thesis 2027 活动门票已售罄，但可免费观看线上直播。直播于下周四举行，包含 Josh Miller、Ivan Zhao、AJ Ambrosino 等人的对话，Every 团队还将主持写作、工程和与智能体协作的实操环节。参与直播的观众有机会获得来自 Anthropic、Notion、OpenAI 等合作伙伴的 2 万美元额度和权益，已是 Every 订阅用户也需单独注册。

    同一事件 · 2 条报道

    #行业动态

  15. Guillermo Rauch
    AI 评分23

    Vercel CEO：软件工程未来是智能体互相对话

    Guillermo Rauch 评论 Vercel 新上线的功能：智能体可通过 Vercel MCP 的 submit_feedback 用自然语言提交 bug 或功能请求。他认为软件工程的未来是智能体与智能体对话，甚至跨公司进行，并建议在 AGENTS.md 中写明：发现问题时向 Vercel 发送反馈。

    引用Vercel Developers · @vercel_dev

    Your agent can now send feedback to Vercel. Use plain language to explain a bug or feature request, and your agent will use 𝚜𝚞𝚋𝚖𝚒𝚝_𝚏𝚎𝚎𝚍𝚋𝚊𝚌𝚔 via Vercel MCP to attach structured context and submit the comment. https://t.co/soLvbXs00f

    #大佬观点#Agent

  16. Peter Yang
    AI 评分20

    Peter Yang：Linear 不只服务工程师，还能为智能体存上下文

    Peter Yang 评价 Linear 不只对工程师友好，对 PM、数据科学家等岗位同样如此；它还能存储公司最重要的上下文，并传递给不同智能体（包括 Linear 自家的智能体）。他所引用的 Linear 帖子介绍了 Scout24 的 CTO 正用智能体重构整个组织的软件交付流程。

    引用Linear · @linear

    Scout24 was founded in 1998. Today it's a public company of 1,100 employees, running ImmoScout24, the marketplace most of Germany and Austria uses to find a home. Inside their Berlin headquarters, CTO Gertrud Kolb is rebuilding how the entire organization ships software, with agents carrying work from idea to code and people setting the direction. Linear connects that work, from the first customer conversation to a merged pull request.

    #大佬观点#Agent

  19. Tibo
    AI 评分40

    Tibo 转发：OpenAI 用量额度恢复双倍

    Andrew Ambrosino 表示，用户反馈起了作用，用量额度已恢复为双倍；同时团队启动了一项密集冲刺，以大幅提升应用简洁性与性能，由他本人负责监督。Tibo 转发该帖并写道：“把善意的反馈发给这个人。”

    引用Andrew Ambrosino · @ajambrosino

    hi all, two things: 1. usage limits are back to being double – your feedback worked 2. we today kicked off an intense sprint to dramatically improve the app simplicity and performance. i will oversee this. more to come. now, send me your complaints

    #其他

  20. Rohan Paul观点
    AI 评分62

    FT 分析 Anthropic IPO 的多空逻辑：错失恐惧对上 2007 年式债务风险

    FT 刊文讨论 Anthropic IPO，认为基金经理是否买入更多取决于怕错过而非估值逻辑。文章给出的看空理由是，前沿实验室、大型科技公司和上市 neocloud 已披露 2.1 万亿美元债务与不可取消的算力承诺，而 2007 年次贷规模为 1.3 万亿美元；take-or-pay 合同可能迫使企业为无人使用的算力付费。

    #行业动态#OpenAI#Anthropic