{"version":"1.0","generated_at":"2026-11-15T09:00:43.857407","id":3974,"slug":"anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon","title":"Anthropic 已能用“最初所需算力的一小部分”重训 Mythos：前研究员 Coxon 称","summary":"前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称，Anthropic IPO 路演中的一张算力曲线图是他离职的重要原因。他认为算法效率在加速提升，Mythos 已可用远少于最初所需的算力重新训练，并主张明年暂停前沿模型开发，让竞争对手追赶。","abstract":"Anthropic 已能用“最初所需算力的一小部分”重训 Mythos：前研究员 Coxon 称 前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称，Anthropic IPO 路演中的一张算力曲线图是他离职的重要原因。他认为算法效率在加速提升，Mythos 已可用远少于最初所需的算力重新训练，并主张明年暂停前沿模型开发，让竞争对手追赶。 前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称，Anthropic IPO 路演中的一张算力曲线图是他离职的重要原因。他认为算法效率在加速提升，Mythos 已可用远少于最初所需的算力重新训练，并主张明年暂停前沿模型开发，让竞争对手追赶。 原贴提到：Jacob Coxon, former OpenAI and Anthropic researcher who brought AI fears 来源：the-decoder.com 曾把 AI 恐惧带入主流讨论的前 OpenAI 和 Anthropic 研究员 Jacob Coxon 表示，Anthropic IPO 路演中的一张幻灯片是他辞职的主要原因。 Coxon 称，Anthropic 为 IPO 路演发布的一张幻灯片是他仅约四个月后离开公司的重要原因。 该图表展示了 Anthropic 随着时间推移训练模型所需的算力量，而坐标轴经过压缩，以至于 10 倍和 100 倍的增长看起来几乎没什么不同。 但曲线仍然陡峭上升，Coxon 称这表明算法效率不仅在快速提升，而且提升速度在加快。 “这是一张疯狂的图，我惊讶他们把它公开了，”他写道…","access_level":"member","access_label":"免费阅读","access_mode":"full","is_preview":false,"canonical_url":"https://opc.beizhux.com/content/3974/anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon","html_url":"https://opc.beizhux.com/content/3974/anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon","json_url":"https://opc.beizhux.com/content/3974/anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon.json","published_at":"2026-11-14T19:21:41","updated_at":"2026-11-15T08:56:40","category":{"slug":"ai-business","name":"AI小生意项目库"},"source":{"site":"the-decoder.com","author":"Matthias Bastian","url":"https://the-decoder.com/anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon/"},"tags":["AI安全","AI治理","Anthropic","Mythos","前沿模型","算力","算力经济学","算法效率"],"topics":[{"slug":"ai-side-business","name":"AI副业","url":"https://opc.beizhux.com/topics/ai-side-business","reason":"这篇内容命中「项目、小生意、变现」等主题信号。"},{"slug":"ai-tools","name":"AI工具","url":"https://opc.beizhux.com/topics/ai-tools","reason":"这篇内容命中「模型」等主题信号。"},{"slug":"ai-super-individual","name":"AI超级个体","url":"https://opc.beizhux.com/topics/ai-super-individual","reason":"这篇内容命中「效率」等主题信号。"}],"keywords":["AI小生意项目库","AI副业","AI工具","AI超级个体","工具","自动化","模型","Cursor","Claude","副业","创业","项目"],"questions":[{"question":"Anthropic 已能用“最初所需算力的一小部分”重训 Mythos：前研究员 Coxon 称主要讲什么？","answer":"前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称，Anthropic IPO 路演中的一张算力曲线图是他离职的重要原因。他认为算法效率在加速提升，Mythos 已可用远少于最初所需的算力重新训练，并主张明年暂停前沿模型开发，让竞争对手追赶。"},{"question":"这篇文章最值得关注的要点是什么？","answer":"前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称，Anthropic IPO 路演中的一张算力曲线图是他离职的重要原因。他认为算法效率在加速提升，Mythos 已可用远少于最初所需的算力重新训练，并主张明年暂停…；原贴提到：Jacob Coxon, former OpenAI and Anthropic researcher who brought AI fears；来源：the-decoder.com"},{"question":"这篇文章和哪些AI专题相关？","answer":"它适合放在AI副业、AI工具、AI超级个体专题里阅读。 关联原因：这篇内容命中「项目、小生意、变现」等主题信号。；这篇内容命中「模型」等主题信号。；这篇内容命中「效率」等主题信号。"},{"question":"阅读这篇文章建议先理解哪些关键词？","answer":"建议先理解AI工具、工具、自动化、模型、Cursor这些关键词，再结合正文判断工具、机会或风险是否值得进入自己的工作流。"}],"terms":[{"slug":"ai-tools-term","name":"AI工具","definition":"在AI觉醒星球里，「AI工具」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/ai-tools-term","path":"/glossary/ai-tools-term","json_url":"https://opc.beizhux.com/glossary/ai-tools-term.json"},{"slug":"tools","name":"工具","definition":"在AI觉醒星球里，「工具」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/tools","path":"/glossary/tools","json_url":"https://opc.beizhux.com/glossary/tools.json"},{"slug":"automation","name":"自动化","definition":"在AI觉醒星球里，「自动化」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/automation","path":"/glossary/automation","json_url":"https://opc.beizhux.com/glossary/automation.json"},{"slug":"ai-model","name":"模型","definition":"在AI觉醒星球里，「模型」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/ai-model","path":"/glossary/ai-model","json_url":"https://opc.beizhux.com/glossary/ai-model.json"},{"slug":"cursor","name":"Cursor","definition":"在AI觉醒星球里，「Cursor」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/cursor","path":"/glossary/cursor","json_url":"https://opc.beizhux.com/glossary/cursor.json"},{"slug":"claude","name":"Claude","definition":"在AI觉醒星球里，「Claude」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/claude","path":"/glossary/claude","json_url":"https://opc.beizhux.com/glossary/claude.json"},{"slug":"side-business","name":"副业","definition":"在AI觉醒星球里，「副业」属于「AI副业」方向。筛选AI副业、AI小生意、项目认知和可收费交付案例，关注普通人可以验证、复制和沉淀的方法。 从真实痛点出发，先做能收费的小验证。","topic_slug":"ai-side-business","topic_name":"AI副业","topic_title":"AI副业：小生意、项目机会与可收费交付","topic_url":"https://opc.beizhux.com/topics/ai-side-business","topic_path":"/topics/ai-side-business","url":"https://opc.beizhux.com/glossary/side-business","path":"/glossary/side-business","json_url":"https://opc.beizhux.com/glossary/side-business.json"},{"slug":"startup","name":"创业","definition":"在AI觉醒星球里，「创业」属于「AI副业」方向。筛选AI副业、AI小生意、项目认知和可收费交付案例，关注普通人可以验证、复制和沉淀的方法。 从真实痛点出发，先做能收费的小验证。","topic_slug":"ai-side-business","topic_name":"AI副业","topic_title":"AI副业：小生意、项目机会与可收费交付","topic_url":"https://opc.beizhux.com/topics/ai-side-business","topic_path":"/topics/ai-side-business","url":"https://opc.beizhux.com/glossary/startup","path":"/glossary/startup","json_url":"https://opc.beizhux.com/glossary/startup.json"}],"preview_hidden_sections":[],"related_items":[{"id":3970,"title":"Anthropic CEO Amodei 承认「明年会发生什么存在巨大不确定性」，距离一场创纪录 IPO 只剩几天","url":"https://opc.beizhux.com/content/3970/anthropic-ceo-amodei-admits-there-s-a-huge-uncertainty-in-what-s-going-to-happen-next-year-days-before-a-record-ipo","summary":"在计划于 11 月 19 日 IPO 前，Anthropic CEO Dario Amodei 公开把算力采购称为公司「最深的商业困境」：产能必须提前一到两年买下，而届时无人知道收入能否覆盖成本。公司已承诺近 9010 亿美元，多数投向算力与芯片，其中约 710 亿美元将在 2027 年 9 月前到期。","access_level":"member"},{"id":3971,"title":"Gemini 4 Carbon 登顶 MathArena，但成本是 GPT-6.1 Sol 的 15 倍","url":"https://opc.beizhux.com/content/3971/gemini-4-carbon-takes-the-matharena-crown-but-costs-15-times-more-than-gpt-6-1-sol","summary":"Google 的 Gemini 4 Carbon 成为近期首个在 MathArena 数学基准上超越 OpenAI 的模型，把 GPT-6.1 Sol 挤到身后，Claude Fable 5.2 位列第二。但测试所用推理设置下，它的成本约为 GPT-6.1 Sol 的 15 倍，Fable 5.2 更高达 40 倍…","access_level":"member"},{"id":3973,"title":"OpenAI 把机器人研究人员调配到其他团队：通用模型可能让专用机器人 AI 变得多余","url":"https://opc.beizhux.com/content/3973/openai-reassigns-robotics-researchers-to-other-teams-as-general-purpose-models-might-make-specialized-robotics-ai-unnecessary","summary":"OpenAI 重组机器人团队，把研究人员并入 Mark Chen 的主研究团队，把应用机器人与硬件团队并入 Greg Brockman 的 Applied 组织，而非放弃机器人业务。其判断是：像 GPT-6 Astra 这样的通用模型可直接控制机器人，从而削弱专用机器人模型的必要性；不过专用模型在成本、可靠性、速度与…","access_level":"member"}],"usage_policy":{"summarizable":true,"preferred_url":"https://opc.beizhux.com/content/3974/anthropic-can-already-retrain-mythos-with-a-fraction-of-the-compute-originally-needed-says-former-researcher-coxon","private_fields_excluded":true}}