标签: DeepSeek

Fascinating interview on how Claude Tag has changed the way work is done at Anthropic 65% of PRs by product & eng teams at Anthropic are now raised by Claude Tag For non-engineering teams, I think the ultimate agent i…

Fascinating interview on how Claude Tag has changed the way work is done at Anthropic 65% of PRs by product & eng teams at Anthropic are now raised by Claude Tag For non-engineering teams, I think the ultimate agent i...

Anthropic 内部访谈显示,其产品和工程团队已有 65% 的 Pull Request(PR)由 AI 编程代理 Claude Tag 提交,同时观察者认为代理交互界面正从终端、桌面应用走向 Slack 等协作工具,AI 正从“开发者的命令行”变成“所有人的工作搭子”。

cool use case of chatgpt work i heard last night: connect your family calendars and explain your kids’ interests. every morning for the drive to school, have it make a podcast that talks about one kid’s soccer game th…

cool use case of chatgpt work i heard last night: connect your family calendars and explain your kids' interests. every morning for the drive to school, have it make a podcast that talks about one kid's soccer game th...

OpenAI CEO Sam Altman分享了一个将ChatGPT用于生成家庭定制播客的使用案例——连接家庭日历、分析孩子兴趣,每天早上为送学车程生成个性化音频内容。但其评论区的高赞回复——“为什么不直接和孩子聊天”——揭示了这个案例真正值得讨论的问题。

No idea if these specific numbers generalize across tasks, but directionally it’s clear that the harness is going to become the most important variable -right next to model capability- in the AI stack. The ability for…

No idea if these specific numbers generalize across tasks, but directionally it’s clear that the harness is going to become the most important variable -right next to model capability- in the AI stack. The ability for...

Box CEO Aaron Levie提出,AI技术栈中“工具链/编排层”(harness)正在成为仅次于模型能力的关键变量。Composio的同日实测数据显示,不同Agent工具链处理同一批任务的平均成本最高相差约3.7倍,成本控制正从模型层转移到调度层。

Teleprompter Pro-X

Teleprompter Pro-X

AI 提词工具 Teleprompter Pro-X 在 Product Hunt 上线,定位面向专业视频创作者的提词与录制场景。目前公开信息显示,产品页尚未完整披露技术细节,但它切入的是短视频和直播内容生产的关键一环。

ScribeRyte AI

ScribeRyte AI

ScribeRyte AI 是一款在 Product Hunt 上线、定位 AI 写作/记录方向的新产品,目前公开信息有限。在同类工具密集的赛道上,它的差异化定位和定价策略值得关注,但不宜过早下结论。

~1500 Elo! Consistently beats frontier models and Stockfish level 0. Fun seeing an 8b model mogging GPT 5.6 with high reasoning and response chaining. Spends 1-2 seconds per move vs 30 seconds. Play it: https://t.co/q…

~1500 Elo! Consistently beats frontier models and Stockfish level 0. Fun seeing an 8b model mogging GPT 5.6 with high reasoning and response chaining. Spends 1-2 seconds per move vs 30 seconds. Play it: https://t.co/q...

有人用 Qwen 系列的 8B 小模型跑出了一个国际象棋 AI,Elo 约 1500,能稳定击败多个前沿大模型和 Stockfish 最低难度,每步棋只需 1-2 秒。这件事的价值在于:小模型在特定任务上,正用远低于旗舰模型的成本做出可用表现。

第 92 期:xAI 联合创始人解读模型开发的未来

第 92 期:xAI 联合创始人解读模型开发的未来

xAI联合创始人在第92期「Unsupervised Learning」播客中分享了对下一代大模型开发的判断,讨论重点从参数规模转向训练效率、推理成本与开源策略。这一表态值得关注,因为它直接关系到大模型竞争的下一步走向和API价格变化的可能方向。