搜索“xAI”

to ensure accuracy, clarity, and responsible discussion of advanced AI concepts. Discover More Baker Lab Designs AI Proteases That Cleave Disease-Linked Proteins; Square Launches Flagship Kiosk With Agentic AI Voice Ordering; xAI Introduces Grok Voice 3.0 Speech-to-Speech Model for Voice Agents; NVIDIA Unveils cuPotential as Semiconductor Leaders Adopt CUDA-X; Microsoft Research Publishes AI Weather Forecasting Study in Nature; Google Rolls Out AI Scam Defenses Across Android Messages and Calls

匹配字段正文

acy, clarity, and responsible coverage of the rapidly evolving AI compute landscape. Discover More Anthropic Adds OpenTelemetry Trace Export to the Claude Platform; Baker Lab Designs AI Proteases That Cleave Disease-Linked Proteins; Square Launches Flagship Kiosk With Agentic AI Voice Ordering; xAI Introduces Grok Voice 3.0 Speech-to-Speech Model for Voice Agents; NVIDIA Unveils cuPotential as Semiconductor Leaders Adopt CUDA-X; Microsoft Research Publishes AI Weather Forecasting Study in Nature

匹配字段正文

considerations” that come with AI. The Department of Defense has previously sided with AI labs, saying in a filing in a lawsuit brought by Memphis communities against Elon Musk’s xAI that the data centers powering AI are essential to military and national security activities. The Latest from Katelyn FBI Arrests OpenAI Employee Suspected of Planting a Surveillance Bug; Can Vibe Coded Apps Replace Adobe and Microsoft? These AI Apps Hope So; You Can Use Gemini 4 Argon Now: Google’s New AI Is Built

匹配字段正文

ra于今年5月被关停，图源：X 但关停不等于放弃回报。此次与初创公司的潜在合作，为Sora提供了一种新的商业化可能：OpenAI通过向初创公司开放技术并换取股权，有机会在不直接运营Sora产品的情况下，继续获得此前研发投入的部分回报。这相当于把Sora从自营产品转为技术授权资产，由外部团队承担运营和定制成本，OpenAI则以股东身份分享潜在收益。 与此同时，xAI和Google继续推进各自的视频模型，Sora留下的市场空间正被对手填补。但是OpenAI并未因此收缩，而是将资源投向其他方向：据萨姆・奥尔特曼（Sam Altman）今年6月1日在X平台发布的招聘信息，OpenAI正式成立机器人团队；另据The Information 9月报道，Patreon联合创始人山姆·杨（Sam Yam）加入OpenAI，领导新设立的“创作者产品”部门，并推进个人设备计划。 结语：Sora复活仍待谈判，其命运何去何从？ Sora能否从停运的消费级产品转型为服务电影制作的专业工具，取决于这场谈判的结果。 若谈判落地，OpenAI不但绕开了算力的成本压力，也借外力把这项技术重新推入专业制作流程。 若谈判失

匹配字段正文

大到 16B 后，样本效率损失超过吞吐收益，登顶反而比 B 更慢。图例给出了归一化 time-to-target。 引言 预训练为大语言模型建立了新的能力基线；强化学习则通过在新环境和新任务中的探索与反馈，决定模型还能继续进步多远。DeepSeek-R1 展示了大规模强化学习如何仅使用少量标注数据发展出强大的推理能力。[^r1] 随着 Grok 4 的发布，xAI 进一步扩大了计算规模：据其介绍，Grok 4 使用拥有 200,000 张 GPU 的 Colossus 集群，以预训练规模运行强化学习，所使用的计算量超过此前 RL 工作的一个数量级。[^grok] 随着强化学习在模型开发中的占比不断提升，其 效率 也会对训练预算产生越来越大的影响。 考虑一个示例预算：假设 10,000 张 GPU 连续运行 30 天，每张 GPU 每小时的成本为 3 美元，那么总成本将达到 2,160 万美元 。如果在达到相同模型质量的情况下，训练时间缩短 10%，那么按照这一假设，可以节省 720,000 个 GPU 小时，即 216 万美元 。对于运行自有集群的团队来说，这些节省下来的 GPU 小时

匹配字段正文

hroughput (blog); [2025/09] SGLang Day 0 Support for DeepSeek-V3.2 with Sparse Attention (blog); [2025/08] SGLang x AMD SF Meetup on 8/22: Hands-on GPU workshop, tech talks by AMD/xAI/SGLang, and networking (Roadmap, Large-scale EP, Highlights, AITER/MoRI, Wave); [2025/08] SGLang provides day-0 support for OpenAI gpt-oss model (instructions); [2025/06] SGLang, the high-performance serving infrastructure powering trillions of tokens daily, has been awarded the third batch of the Open Source AI Gr

匹配字段正文

maticians are creating an AI research alliance led by Fields medalist Shing-Tung Yau, with Yau pushing back against backlash toward the tech following OpenAI’s 700+ paper release. xAI introduced a feature for Grok Bot that allows the agent to obtain and run its @mail.grokbot email address, letting it email businesses and book meetings for users. Anthropic unveiled the Anthropic Cyber Mission, an initiative to utilize Claude for cybersecurity defense for critical infrastructure and open-source co

匹配字段正文

蒂亚诺，今年9月正式加入了OpenAI基金会董事会。制定评估标准的人，坐进了被评估公司的治理层，“既当运动员又当裁判员”的争议随之而来。 “以安全为名协同减速”是否越过了分界线，如今已不止是猜测：9月18日，四名分别使用ChatGPT、Claude、Grok、Gemini的付费用户在美国加州北区联邦地方法院提起集体诉讼，指控Anthropic、OpenAI、xAI与谷歌达成非法协议、联手放缓研发，涉嫌违反反垄断法。 另一方面，在达里奥、山姆·奥特曼、马斯克呼吁“暂停”前不久，他们各自掌舵的Anthropic、OpenAI、Grok也刚刚发布过最新模型，且大多出了降价举措，如Claude Opus 5.5单价直降20%，Anthropic在宣传中更表示典型任务实际成本降约40%；GPT-6 Sol的价格相比上一代几乎打了五折——巨头本身也在悄悄搞促销。 真相可能就藏在夹缝里：事故是真的，所以刹车是必要的；但手握刹车的，恰恰也是最想独占赛道、最擅长把“安全”变成准入门槛的那几家。真正的风险预警，必须落在对所有人一视同仁的规则里——而不是某些公司既当运动员，又当裁判，还负责给其他车手划线。

匹配字段正文

del-Cards/Gemini-3-Pro-Model-Card.pdf; Gemini 3.1 Pro. Google DeepMind. (2026). Gemini 3.1 Pro model card. https://deepmind.google/models/model-cards/gemini-3-1-pro/; Grok 4 Fast. xAI. (2025). Grok 4 Fast model card. https://data.x.ai/2025-09-19-grok-4-fast-model-card.pdf; DeepSeek-V3. DeepSeek-AI. (2024). DeepSeek-V3 technical report. arXiv. https://doi.org/10.48550/arXiv.2412.19437; DeepSeek-R1. DeepSeek-AI. (2025). DeepSeek-R1: Incentivizing reasoning capability in LLMs via reinforcement lear

匹配字段正文

d problems [link] Karl Cobbe, et al., 2021. arXiv preprint arXiv:2110.14168; Introducing GPT-5 [link] OpenAI, Aug. 7, 2025; Hello GPT-4o [link] OpenAI, May 13, 2024; Grok 4 [link] xAI, July 9, 2025 Acknowledgements We thank Aditya Singh and Julian Minder for valuable discussions that shaped this project. We are also grateful to Tim Hua, Andy Arditi, and Clément Dumas for their ongoing feedback throughout this work. We additionally thank the Supervised Program for Alignment Research (SPAR), which

匹配字段正文

sumers, according to a16z. Now startup agents like Instinct and Tomo report hundreds of thousands of users, and the big players are countering with Meta's Muse, OpenAI's Dots, and xAI's Grok Bot. Agents become an economic force once they start triggering real purchases. Instinct founder Noah Shinn says 40 percent of users link a credit card within three weeks and then spend four-figure amounts per month through the service, much of it on travel. As revenue grows, platforms are picking sides. Ama

匹配字段正文

案不再显而易见。 01 过去两年，AI竞争很容易被讲成一场模型竞赛，谁的模型更聪明，谁就赢得用户。这套叙事放到2026年仍然说得通，但解释力在流失。 斯坦福《2026人工智能指数报告》指出，过去一年前沿模型在Humanity's Last Exam（HLE）测试上的成绩提升约30个百分点；截至2026年3月，LMArena模型评估排行榜上Anthropic、xAI、Google、OpenAI四家公司的得分已挤在约25个Elo点以内，美国与中国头部模型的性能差距也缩小到个位数。价格变化进一步强化了同一趋势。OECD 2026年7月发布的AI市场研究显示，2024年1月至2026年4月，文本生成模型的质量调整价格指数下降了近80%，编码、Agent等更复杂能力的降价同样明显。几乎同一时间，模型与外部工具之间的连接也在标准化——MCP（Model Context Protocol）等开放协议像一枚“统一插头”，让AI得以用相对标准的方式连接外部工具、数据和系统，模型本身的可替换性随之增强。 模型正在从稀缺资产变成基础能力，按需采购、随时调用、低成本替换。大模型评测机构Artificial A

匹配字段正文

for any given task, including Claude Opus 5.5, MidJourney, Suno and other leading APIs. Whatever is most likely to give you the best outcome.” Grok Bot is a joint product of SpaceXAI, the Musk AI lab now folded into SpaceX, and Cursor, which SpaceX agreed in June to buy for $60 billion in stock, a deal that closed in August. In August, I wrote that when models converge on price, the money moves to whoever decides which model gets the job. This trend is happening across organizations increasingl

匹配字段正文

ranking 10 of 29 models, behind Qwen-Audio-3.1-TTS-Plus at 84.6% and ahead of Inworld Realtime TTS-2 at 80.7%. It ranks 2 in Preserving Exact Sequences at 83.3%, behind only SpaceXAI TTS. ➤ Pricing: HeyGen Voice is priced at $30/1M characters, compared to Qwen-Audio-3.1-TTS-Plus at $19.3/1M characters, Eleven v4 Turbo at $40/1M characters and Eleven v4 at $80/1M characters. ➤ Speed: HeyGen Voice processes 40 characters per second of generation time, compared to 76 characters per second for Elev

匹配字段正文

on parameters. With a 1-million-token context window, native vision, and built-in prompt caching, it helps you build smarter apps 2.5x faster at lower cost than previous versions. xAI’s frontier models Grok 4.6 and Grok 4.7 offer a 500K token context window and support configurable reasoning effort and self-verification for more reliable long-running tasks. Together, these additions give you more flexibility when matching models to your requirements such as context length, latency, throughput, a

匹配字段正文

ENEMY!” The same tech company figures who rushed to donate millions to Trump after his election have quickly signed on. Trump’s “First Buddy” Elon Musk has promised to rename SpaceXAI to SpaceXSI, while Jeff Bezos declared he likes it because “artificial is not a very flattering thing.” Jensen Huang, whom Trump called live on stage last month to rant that “robots will not be taking over,” has likewise adopted the term. Sam Altman praised it too, though he said OpenAI won’t be changing its name.

匹配字段正文

to better account for the full set of factors lawyers value, including usability features like style and tone. Key takeaways: ➤ Top of the leaderboard: Grok 4.7 (xhigh) from @SpaceXAI leads at 9.4% Hallucination-Gated All-Pass Rate, narrowly ahead of Muse Spark 1.3 (max) from @AIatMeta at 8.9% and GPT-6 Astra (max) from @OpenAI at 8.6% ➤ Hallucinations reshape the leaderboard: without the hallucination gate, Muse Spark 1.3 (max) would lead clearly with a 26.7% all-pass rate, but two thirds of th

匹配字段正文

respondents are saying: AI spend has doubled year on year; 90% expect AI to generate new work; Only 5% have all agent controls enterprise-wide Check out the full survey here. SPACEXAI 🔀 Grok Bot opens up routing to Claude, other models Image source: xAI / Anthropic The Rundown: Elon Musk announced that SpaceXAI’s Grok Bot agent platform will send tasks to rival models like Anthropic’s Claude and image maker Midjourney, with Musk saying it will route to “whatever is most likely to give you the be

匹配字段正文

nthusiasts have been using agents for a minute now — homebrew OpenClaw setups led to a surge in Mac Mini sales earlier this year. But the launch of Meta’s Muse, OpenAI’s Dots, and xAI’s Grok bot has brought easy to use agents to millions. Muse and Dots have had the highest-profile product launches, and they’re fascinating to pit against each other. Both Meta and OpenAI have decided to pitch these agents to mainstream users and businesses in the form of cute, animated mascots . These can do every

匹配字段正文

ctually improving engineering work; Feedback and improvement. A factory needs to learn from what happened rather than repeat the same mistakes. Shivam Jindal's session on how SpaceXAI uses Cursor gives us one view of what that can look like in practice, with long-running agents working through investigation, review, CI failures and low-risk changes The interesting part is how these layers reinforce one another. Better context lets agents do more useful work, which increases the pressure on orche

匹配字段正文