AI News 2026-08-25 — Top 10 AI Stories
Curated AI news digest for 2026-08-25 (timezone Asia/Shanghai).
- OpenAI is building AI agents for everything. Will everyone use them? — TechCrunch AI. Inside the frontier lab’s push to bring AI agents from software engineers to the masses.
- Advancing price-performance for developers with GPT‑5.6 in Kiro — OpenAI News. GPT‑5.6 is now available in Kiro, helping developers plan, build, review, and test software with better price-performance.
- 阿里视频大模型Wan3.0正式上线,行业评价“稳定、真实、有质感” — 量子位. 8月24日,阿里巴巴视频生成大模型Wan3.0正式上线。
- Hugging Face reportedly in talks to be acquired for $13B — TechCrunch AI. Hugging Face has reportedly been fielding acquisition offers that would value the company at around $13B. But with the founders' feeling of responsibility to community, doubts arise as to whether a sale will happen.
- 为什么手机内存进入英伟达机柜后,贵过HBM? — 雷锋网. 过去长期由HBM主导的AI服务器内存账单,正在被“手机内存”改写。 8月11日,外媒Wccftech援引美国银行(BofA)测算称,英伟达面向Vera Rubin Ultra NVL144的Kyber机柜将配置124.4TB HBM4E,成本约为250万美元;同柜216TB LPDDR5X虽然单价更低,但总成本趋近280万美元,超出前者约12%。 LPDDR5X原本主要用于手机等移动设备,例如iPhone 17 Pro搭载12GB LPDDR5X内存。如果做一个换算,Kyber单柜所使用的相当于约1.8万部手机的运行内存总量。 这并不是LPDDR第一次…
- 阿里拟配售800亿港元新股,全部投入AI建设;网友造谣「小红书因遭举报上市失败」被行拘;李斌:中国汽车行业三五年后玩家基本确定 — 雷锋网. 要闻提示 1.“我们大学硕士不就白读了吗”,上市公司被曝要求应届生离职与“打螺丝”二选一 2.阿里巴巴拟配售800亿港元新股,全部投入AI建设 3.网友造谣“小红书因遭举报上市失败”被行拘 4.李斌:中国汽车行业三五年后玩家基本确定 5.雷军:十车PK,小米SU7 Pro续航CLTC排第四、实测排第一 6.马斯克母亲谈儿子推荐大家来中国:埃隆发自内心喜爱中国,中国人非常智慧 7.董宇辉等主播走后东方甄选扭亏为盈:从亏1.10亿到盈利6.63亿 8.外媒:英伟达搭载人工智能芯片的服务器价格将上涨超15% 今日头条 “我们大学硕士不就白读了吗”,上市公司被…
- Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research — TechCrunch AI. Built by DeepMind alumni, British AI lab Inherent released Faraday, an AI agent whose ability to replicate scientific papers could be a stepping stone for innovation.
- ResNet作者任少卿机器人创业!公司注册就独角兽了 — 量子位. 虽然明确机器人创业,但依然在职蔚来
- They Dedicated Their Lives to Teaching. Then the Deepfakes Started — WIRED AI. The deepfake epidemic in schools is affecting more than students. Four teachers tell WIRED about becoming targets of sexualized, AI-generated content—and how difficult it was to find accountability.
- Kids outlearn AI—and we still don’t know why — MIT Technology Review AI. People have been talking to each other for at least 100,000 years, as best we can tell. And in all that time, there has been only one thing in the world that could learn a human language to perfect fluency: a human child. Now there are two. Four short years after the release of…
Back to latest AI News