AI 每日快讯
AI 产品、模型、开源工具和官方动态的时间流。保留历史记录，按分类、日期和标签继续筛选。
一句话结论：How Jump Trading is scaling quant research with ChatGPT。原始信息显示，这条动态来自 OpenAI News，摘要线索为“Jump Trading uses OpenAI to expand quantitative research. See how longer-running AI 工作流 combine multiple data sources with human review.”。它属于“Agent 与工作流”方向，可能影响工具选择、内容生产、开发工作流或企业落地判断。适合 AI 工具用户、运营/产品团队、开发者和正在做 AI 选型的中文用户阅读。下一步建议先核对原文，看它解决的具体任务、适用范围、价格或部署条件，再对照站内工具库、专题和知识图谱，判断是否值得收藏、试用或写成教程。
原文摘要：OpenAI publishes new results on open problems in mathematics from an internal frontier model and shares Lean proof formalizations and research details on GitHub. 来源：OpenAI 官方动态。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Google released EmbeddingGemma 2, an open model with 740 million parameters that converts text, images, video, audio, and code into vectors. It runs on-device, needs only 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Google's new Nano Banana 2.1 image model uses Gemini 3.6 Flash and beats the previous Pro model in some 评测 at a lower cost. But its predecessor also scored well in 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Google DeepMind's EmbeddingGemma 2 maps 5 input types into one 768d space and ships today under Apache 2.0. The post Google DeepMind Releases EmbeddingGemma 2, a 740M Open Multimod 来源：MarkTechPost。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：According to the Wikimedia Foundation, rogue OpenAI agents edited wikis without permission, tried to abuse a citation tool as a proxy, and may have caused a partial Wikida 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
一句话结论：Advancing computer use with Ironclad。原始信息显示，这条动态来自 OpenAI News，摘要线索为“Learn how OpenAI and Ironclad are training and evaluating AI agents on complex contracting 工作流 to advance computer use for professional work.”。它属于“Agent 与工作流”方向，可能影响工具选择、内容生产、开发工作流或企业落地判断。适合 AI 工具用户、运营/产品团队、开发者和正在做 AI 选型的中文用户阅读。下一步建议先核对原文，看它解决的具体任务、适用范围、价格或部署条件，再对照站内工具库、专题和知识图谱，判断是否值得收藏、试用或写成教程。
原文摘要：Microsoft published a bearish AI outlook from Nobel economist Daron Acemoglu. He predicts about 1.5 percent GDP growth over ten years and at most five percent of jobs repl 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Mistral AI has released Mistral Large 4, nicknamed Le Chonk, as a public preview. It is a 1.05 trillion parameter Mixture of Experts model with 49 billion active parameters, native 来源：MarkTechPost。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Atlassian and OpenAI are expanding their partnership to connect frontier models with enterprise knowledge and help teams plan, build, and deliver work. 来源：OpenAI 官方动态。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Insurers are bracing for millions in claims from rogue AI agents, and executives like OpenAI's Sam Altman and Anthropic's Dario Amodei could be personally on the hook for 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Mistral's Large 4 is the company's biggest model yet, with one trillion parameters trained on its own European infrastructure. In the independent Intelligence Index, the m 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：South Korea wants to develop a homegrown frontier model through government-backed equity investments of 4.7 trillion won ($3.49 billion). The article South Korea bets $3.4 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Telecom operators are increasingly building their AI strategies on open models — and the reasons go beyond mere cost. Open models give telcos the ability to trust, control and cus 来源：NVIDIA AI 动态。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Reflection has released Beam, its first open-weight model. The mixture-of-experts system activates just 23 billion of its 501 billion parameters per token and aims to matc 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Researchers at PhAI Labs have expanded Yann LeCun's JEPA architecture to work across seven fields, from robotics to biomedicine. The effort also produced a liver cancer tr 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Deepseek is close to raising at least $12 billion in a new funding round, Bloomberg reports. The article CATL and Tencent back Deepseek's ballooning funding round as the A 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Reka has released Rho-1, a 19B omni-reasoning model trained from scratch. One network reads and generates text, images, video and robot actions over a shared KV cache. A distilled 来源：MarkTechPost。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：Cohere turned its enterprise platform into a control center for AI agents with North 2, handling multi-step 工作流 on their own and retaining context across sessions. T 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
原文摘要：JEPA-Anything splits a JEPA's single latent target into 4 orthogonal factors, each with its own predictor. Tested across 7 domains, it beat matched JEPA baselines on all 10 dynamic 来源：MarkTechPost。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。