生成式AI

Search documents
腾讯研究院AI速递 20250529
腾讯研究院· 2025-05-28 15:06
生成式AI 一、 576亿!云 计 算 与SaaS 巨头 Informatica 收购 Informatica 1. Salesforce以80亿美元收购云数据管理龙头Informatica,为其自2021年收购Slack后最 大交易; 2. 此次收购将整合双方AI引擎,打造可信数据基础架构,支持代理式AI系统的企业级部署; 3. 数据管理能力成为企业AI产品的关键差异化优势,Salesforce通过收购完善其数据管理领 域布局。 https://mp.weixin.qq.com/s/2knNvn93HIE4OVEeeqF9LA 二、 DeepSeek 官方推送通知,称 R1 模型已经完成小版本试升级 1. DeepSeek R1模型完成小版本升级,现已可在官方网站、APP和小程序中体验; 2. 实测显示R1编程能力有明显增强,能快速生成高质量的动态天气卡片,包含细致设计和交 互动画; 3. 此次更新可能采用了DeepSeek-V3-0324(685B参数)模型,而备受期待的R2版本仍未 发布。 https://mp.weixin.qq.com/s/rJAqhBclKlswtV24CDFKEg 三、 Anth ...
《星球大战》出品人现身成都:这里有种“科幻电影”即视感
Mei Ri Jing Ji Xin Wen· 2025-05-28 14:09
每经记者|石普宁 每经编辑|唐元 图片来源:每经记者 张建 摄 "成都的城市建筑极具视觉冲击力,本身就像一部科幻电影中的场景。"5月28日,在成都天府长岛数字文创园——这个孕育了现象级动画电影《哪吒》的地 方,D'Art Shtajio动画公司创始人阿特尔·伊索姆(Arthell Isom)在接受《每日经济新闻》记者专访时,分享了他对动漫、AI技术与成都的看法和印象。 当天,"全球联动·产业互动"——科幻产业国际沙龙在此举行。活动由每日经济新闻主办,每经智库承办,成都市对外开放合作促进中心、成都市法学会、 游戏茶馆、共佳品创支持。 室外微风习习,即便是一身西装,也不难看到阿特尔·伊索姆练得一身好肌肉。当记者称赞其身材时,他笑着回应:"搞创作的一定要把身体搞好。"在这样 轻松的氛围中,我们的对话就此展开。 D'Art Shtajio是日本首家有美国背景、也是动漫史上首家由非裔主营的大型动漫工作室,曾为《进击的巨人》《海贼王》《东京食尸鬼》等多部热门作品提 供过制作服务。 阿特尔·伊索姆则是《星球大战:幻境》(Star Wars: Visions)第一季的出品人,并深度参与过《进击的巨人》《海贼王》《银魂》《柯南 ...
假新闻泛滥,谁杀死了可信信息?
Hu Xiu· 2025-05-28 12:21
2023 年底,新闻可信度评级机构 NewsGuard 发布了一项数据,显示在 2023 年 5 月到 12 月这个时间段,由人工智能驱动的假新闻网站增长了 1000%,数 量从 49 个激增到 600 多个。 而到 2024 年 6 月,这个数字已经变成 1265 个。NewsGuard 将其称为"一个可悲的里程碑"。因为根据测算,人工智能驱动的假新闻网站,数量正式超过了 美国本土的报纸网站数量(1213 家)。 这确实是新闻业的可悲现状。传统的、值得信赖的、机构化的本地新闻来源正在大批次地湮灭,2005 年到 2022 年期间,美国约有 2500 家当地报纸倒 闭,取而代之的却是大量的 AI 假新闻网站。这些网站正成为许多不明真相者的信息来源,并进一步流传到社交网络中,变成流行甚广的虚假社交货币。 AI 假新闻正在泛滥,这样的趋势同样在国内发生。今年年初,西藏日喀则定日县发生 6.8 级地震,多张"小男孩被埋图"在网络疯传,引发广泛关注。后经 公安机关核查,这些图片均由一位青海网友使用 AI 工具生成,并上传网络,引发谣言扩散。 新闻截图 不久前"顶流男星豪赌 10 亿"也是相当典型的案例。根据公安部的通 ...
中国扫货半导体设备
半导体芯闻· 2025-05-28 10:17
如果您希望可以时常见面,欢迎标星收藏哦~ 来源:内容编译自theelec 。 去年,中国大陆在购买用于半导体生产的晶圆厂设备上的支出是各个国家和地区中最高的。 据SEMI统计,全球第二大经济体中国大陆在晶圆厂设备上的支出为495.5亿美元,同比增长35%。 | Region | 2024 | 2023 | (YoY) % | | --- | --- | --- | --- | | China | $49.55 | $36.60 | 35% | | Korea | $20.47 | $19.94 | 3% | | Taiwan | $16.56 | $19.62 | -16% | | North America | $13.69 | $12.05 | 14% | | Japan | $7.83 | $7.93 | -1% | | Europe | $4.85 | $6.46 | -25% | | Rest of the World | $4.19 | $3.65 | 15% | | Total | $117.14 | $106.25 | 10% | SEMI指出,中国凭借积极的产能扩张和政府扶持国内半导体产业的 ...
智算业务能救「独立云厂商」吗?
雷峰网· 2025-05-28 07:52
" 始于卖资源的智算生意,搏的是当下的资本,还是对未来的确 信? " 作者丨赵之齐 编辑丨胡敏 "未来智算业务一定会成为独立云厂营收的主要支柱,不然也没有别的支柱了",算力行业人士王志苦笑。 2022年底,王志所在的独立云厂,也在卖CDN、公有云价格战中伤痕累累。当时云计算市场头部集中, 留给中小厂商的空间微乎其微;即便是成功上市的优刻得(UCloud)、青云科技、金山云等,持续的营 收下滑和亏损加剧,也已让资本们有些失去信心。 直到生成式AI的出现,独立云厂窥见新的希望,王志所在的公司开始了在智算领域的第一个尝试:卖资 源。他清晰地记得,2023年初完成了第一单A800机器成交后,脑中浮现出的第一个念头: "赌对了!" 当时,一台八卡的A800机器租赁价七八万一个月,市场已觉得有些昂贵,谁知,在那之后两年,算力价 格一度水涨船高——2023年底,英伟达推出的包括A800、A100、H800等五种特供显卡被列入禁令,加 上2024年的AI狂奔,有行业人士透露,A800服务器市价曾达到100-200万一台,H800服务器达到200- 300万一台,A100价格也从10万炒到了20万。而王志所在的公司,智算业务布 ...
TechNet中国2025:商汤科技(0020.HK)推出基础模型;拓展AI驱动的用户案例
Goldman Sachs· 2025-05-28 05:15
27 May 2025 | 7:19AM HKT TechNet China 2025: SenseTime (0020.HK) Foundation model introduced; expanding AI-powered user case We hosted SenseTime's management on May 21 at our TechNet Conference China 2025 in Shanghai. Management remains positive on the generative AI trend in China, and highlights their newly launched foundation model, SenseNova V6, carrying upgraded features with competitive costs across training and inferencing. The company also newly signed a MOU with the Faculty of Law at the Chinese Uni ...
华为正式推出昇腾超节点技术,资金连续8日净流入场内规模最大的计算机ETF(159998)
2 1 Shi Ji Jing Ji Bao Dao· 2025-05-28 03:01
Group 1 - A-shares indices opened higher on May 28, with the Computer ETF (159998) experiencing a slight decline of 0.47% and a trading volume exceeding 20 million yuan [1] - The Computer ETF (159998) has seen continuous capital inflow over the past 8 trading days, accumulating a net inflow of 112 million yuan, making it the top performer in its category [1] - The latest scale of the Computer ETF is 2.801 billion yuan, making it the largest computer ETF by on-market size [1] Group 2 - The Cloud Computing ETF (517390) has recorded net inflows on 4 out of the last 5 trading days, indicating strong investor interest [2] - The Computer ETF tracks the CSI Computer Index, which includes stocks from companies involved in information technology services, application software, system software, and computer hardware, with top holdings including Hikvision, iFlytek, Kingsoft Office, and others [2] - Huawei has launched the Ascend Super Node technology, which consists of 12 computing cabinets and 4 bus cabinets, achieving the industry's largest scale of 384-card high-speed bus interconnection [2] Group 3 - The AI sector is undergoing rapid evolution, shifting focus from model scale and benchmark performance to user experience and interaction innovation, leading to a new round of industry reshuffling [3] - Generative AI is transitioning from passive response to active execution of complex tasks, enhancing its practical application scenarios and commercialization potential [3] - Domestic AI computing power ecosystems, represented by Ascend, are improving through innovations in underlying architecture and tools, boosting the performance of related industries such as high-speed connectors and liquid cooling [3] Group 4 - As general reasoning capabilities advance, high-value applications in research and programming are expected to unlock first, benefiting software and internet sectors [3] - Hardware demand is anticipated to rise alongside advancements in multimodal technology, maintaining a positive outlook on investment opportunities in the AI computing power sector [3]
510万股,苹果突遭抛售!年初以来股价跌超20%,市值蒸发超5.5万亿元!发生了什么?
Mei Ri Jing Ji Xin Wen· 2025-05-28 02:55
每经编辑|毕陆名 近日,美国最大公共养老基金——加州公务员退休基金(CalPERS)对其投资组合进行了重大调整,大幅减持了苹果公司的股票。 美国证券交易委员会(SEC)披露的文件显示,一季度,CalPERS抛售了510万股苹果公司股票,持仓规模降至3470万股。与此同时,CalPERS增持了 Meta、AMD和麦当劳的股票,增持数量分别为57.9万股、32.5万股和49.4万股。 根据公开资料,CalPERS管理超过5400亿美元(约合人民币近4万亿元)的资产,以支持200多万加州公务员的退休福利。该养老基金一季度抛售苹果股票的 举动,可能预示着大型机构投资者对苹果的信心下降,也加剧了市场对苹果前景的担忧。 据证券时报,对于抛售苹果股票的报道,CalPERS回应媒体称:"公司的公共资产投资是以指数为导向的,并使用系统和定量的投资策略进行优化,不受任 何单一时期事件的驱动。因此,我们通常不会对我们的个人持股或交易发表评论。我们的团队将继续评估整体市场状况。" 苹果公司的业务及股价受关税影响比较大。该公司5月初预计,将在第三财季(2025年3月30日至6月28日)因美国关税政策损失约9亿美元。另外,美国总统 特朗 ...
腾讯研究院AI速递 20250528
腾讯研究院· 2025-05-27 15:44
Group 1 - UAE becomes the first country to offer free access to ChatGPT Plus for all citizens, part of a collaboration with OpenAI [1] - Abu Dhabi will establish the Stargate UAE high-performance AI data center, supporting a 1 GW computing cluster with an initial target of 200 MW capacity [1] - The collaboration is part of OpenAI's "nation-focused" initiative, with UAE committing to match US funding, potentially totaling up to $20 billion [1] Group 2 - OpenAI has enabled singing capabilities for GPT-4o, seen as a response to Google's Gemini 2.5 Pro and Veo3 releases [2] - Google's Gemini 2.5 Pro has outperformed OpenAI and Claude models in several benchmark tests [2] - Analysts believe that the singing feature of GPT-4o is insufficient to regain market leadership, emphasizing the need for OpenAI to launch GPT-5 soon [2] Group 3 - Claude Opus successfully solved a stubborn bug that had troubled a veteran C++ engineer for four years, taking only a few hours [3] - The AI identified the root cause of the issue through analysis of code libraries and architecture comparisons, which had previously stumped other models [3] - Despite its debugging prowess, AI is still considered to be at a beginner level in writing new code [3] Group 4 - French non-profit AI research organization Kyutai launched Unmute, a modular voice AI system that can quickly add voice interaction capabilities to any text LLM [4] - Unmute features low latency (200-350 ms), streaming speech-to-text and text-to-speech, full-duplex interaction, and 10-second voice cloning, supporting over 70 emotional styles [5] - Kyutai plans to fully open-source Unmute in the coming weeks, including STT (1B parameters) and TTS (2B parameters) models and code [5] Group 5 - Alibaba Tongyi launched QwenLong-L1-32B, a large model addressing long-context reasoning issues, with a maximum context length of 130,000 tokens [6] - The team identified two core challenges: low training efficiency and instability, proposing progressive context expansion techniques and a mixed reward mechanism [6] - QwenLong-L1-32B outperforms models like OpenAI-o3-mini and Qwen3-235B-A22B, showing significant advantages in long document analysis [6] Group 6 - Mita AI Search introduced a new "Ultra" model, achieving a response speed of 400 tokens per second, with most queries answered within 2 seconds [7] - The new model utilizes kernel fusion on GPUs and dynamic compilation optimization on CPUs, achieving performance breakthroughs on a single H800 GPU [7] - Mita offers both "Ultra" and "Ultra·Thinking" modes optimized for different types of questions, along with a temporary speed test site for user experience [7] Group 7 - Thunderbird officially released the AI glasses X3 Pro, featuring a custom large model and full-color display, priced at 8,999 yuan [8] - The X3 Pro utilizes a 4nm Qualcomm Snapdragon AR1 platform and proprietary Firefly light engine with RayNeo waveguide technology, achieving a brightness of 3,500 nits (peak 6,000 nits) and weighing only 76g [8] - The product is available for pre-order and will ship on June 15, supporting AI Agent store and real-world navigation features [8] Group 8 - The core team of Meta's Llama faces significant talent loss, with 11 out of 14 core authors having left, leaving only 3 remaining [10] - Among the departed, 5 joined the French AI open-source startup Mistral, including two main architects of Llama [10] - Meta is under pressure from open-source models like DeepSeek and Qwen, despite investing billions, lacking a dedicated "inference" model [10] Group 9 - The Beihang University team proposed the "Flying-on-a-Word" (Flow) task, enabling drone control through language commands, filling a gap in low-level language interaction control research [11] - The team constructed the UAV-Flow benchmark dataset, containing 30,000 real-world flight trajectories across eight major movement types [11] - The research addressed drone computational limitations by performing model inference at the ground station and providing real-time feedback for control commands [11] Group 10 - NVIDIA experts recommend that students integrate multiple skills and enhance adaptability, not limited to computer science backgrounds, to stand out in the job market [12] - Job seekers should clarify their interests in the AI field, responsibly use AI tools, and build industry connections for career development opportunities [12] - Candidates can showcase their technical abilities, professional knowledge, and innovative thinking through project examples to excel in interviews [12]
硅谷最疯CEO:卖掉摇钱树《宝可梦GO》后,他做了什么?
创业邦· 2025-05-27 10:11
以下文章来源于福布斯 ,作者Forbes 福布斯 . 福布斯中国集团是一家集咨询、社群于一体的企业。集团秉承企业家精神和创新精神,坚持专业、公 正、创新和进取的价值观。 作者丨 Richard Nieva 翻译丨 Lemin 图源丨 Nikkei Asia 走进旧金山历史悠久的渡轮大厦内的 Niantic 总部,访客首先会被一群巨大的宝可梦毛绒玩具包围: 在阶梯式剧场般的台阶上,一只巨大的卡比兽正在角落打盹,而妙蛙种子则仿佛随时准备扑向猎物。 不远处,一脸震惊的可达鸭茫然望着远方,似乎想要看清这家公司出人意料的未来。 来源丨福布斯(ID: forbes_china ) 今年 3 月, Niantic 宣布了一则重磅消息: 根据高德纳( Gartner )的数据, 2023 年空间计算的市场规模为 1100 亿美元,到 2033 年预计将 增至 1.7 万亿美元,增长动力来自 TomTom 等地图巨头以及谷歌等传统科技巨头的定位服务。 "这是 个巨大的机会。"高德纳新兴技术团队首席分析师阮祥( Tuong Nguyen ,音译)表示。 但竞争也同样激烈。在空间 AI 领域, Niantic 面临一些强劲对手。 ...