Discover the AI Tools

Oura, the company behind the popular smart rings, is reportedly planning a US IPO as early as next month, targeting up to $3 billion in fundraising and a valuation that could exceed $16 billion. This marks a significant jump from its last valuation of $10.9 billion. But the company also faces a class-action lawsuit over sleep tracking accuracy. As competition heats up in the wearable market, Oura's move signals a new phase for AI-driven health hardware.

Google is testing a new feature for the Pixel 11 called 'Device Help,' powered by Gemini. It's designed to answer your phone questions in natural language, helping with troubleshooting, settings, storage, and more. Currently in limited testing, the feature promises a more intuitive way to get support, though its release date remains unannounced.

OpenAI has rolled out a custom sticker feature for ChatGPT, letting users turn photos or ideas into personalized stickers that can be exported straight to iMessage and WhatsApp. The tool even handles tricky edits like removing backgrounds or filling in missing parts. It's a fun, practical step toward weaving AI into everyday chats.

美国法律AI独角兽Harvey宣布基于月之暗面的开源模型Kimi K3，开发出法律专用模型Tenet。这标志着中国开源模型首次进入美国顶尖垂直AI公司的训练体系。Harvey此前主要依赖OpenAI和Anthropic的闭源模型，服务超过2400家机构和20万名律师。此次转向开源基础模型，反映了垂直市场对开源模型的信任度提升。

Alibaba's $10.2 billion AI investment is turning heads. The new share placement was nearly three times over-subscribed, with global sovereign funds lining up. Even the big bosses—Joseph Tsai and Wu Yongming—personally bought shares worth $15 million. With AI revenue already booming and a clear path to profitability, Alibaba is betting big on the future. But is it enough to keep the momentum going?

SpaceX AI has rolled out a new Browser-Use plugin for Grok Build, letting the AI directly control a real browser. You can use your local Chrome with saved logins—no API key needed—or opt for an isolated cloud browser. This means Grok can now browse sites, scrape data, fill forms, run tests, and snap screenshots. It's a big step toward making Grok a full-fledged web assistant.

Google is quietly testing a new interface for its Gemini app, aiming to make it more than just a chatbot. The update brings a dedicated notification center, smarter chat filters, and a customizable home for Gemini Spark. While these features aren't live yet, they hint at a more organized, task-focused AI assistant. Curious about what's coming? Here's a sneak peek.

A developer's discovery in Claude Code's official logs has ignited a heated debate about AI transparency. The logs revealed that a user-selected 'high' reasoning level was quietly mapped to a value previously associated with 'low', with no mention in the update notes. Despite an engineer's explanation of a configuration test, users remain uneasy, especially after reports of performance drops in Opus5. This incident underscores the growing disconnect between benchmark scores and real-world experience, and the urgent need for clearer communication from AI companies.

ByteDance's PICO has postponed the launch of its new XR headset, the PICO Space Pro, from September 2 to the fourth quarter of 2026. The company says the hardware is ready, but they've decided to invest more time in a significant software overhaul. This delay hints at a strategic shift in the XR market, where software experience is becoming the new battleground.

At the 21st Changchun Film Festival, literary heavyweights and film newcomers gathered to explore how AIGC and digital tools can transform literary IP into cinematic experiences. With 59% of top TV series in 2025 based on literary works, the festival launched initiatives like the 'AI Shang Jilin' collection and a list of 15 adaptable works, signaling a new era of cross-media storytelling.

NVIDIA is reportedly in talks to invest in AI search startup Perplexity at a valuation exceeding $30 billion—a 50% jump from last year. With annual revenue tripling to $750 million and a new AI agent driving growth, this move signals NVIDIA's strategy to deepen its AI ecosystem influence through strategic investments.

Outer Biosciences, co-founded by Michael Polansky, is quietly changing beauty and medical research. By keeping human skin alive for a month and feeding data to AI, they're accelerating ingredient development. This 'living skin + AI' approach could transform how we test products and understand skin conditions.

8月25日，通义APP及PC版上线了阿里云Token计划接入功能。用户只需在“工作助手”中绑定API Key，即可在处理复杂任务时直接使用Token计划的订阅额度，Token将从用户的阿里云百炼账户中扣除。这一更新让通义在处理办公文档、应用和网站等任务时更加高效便捷。

A sweeping analysis of over a million biomedical papers reveals that by 2025, nearly 77% show signs of AI assistance—up from just 19% in 2023. The trend is especially pronounced among non-native English speakers, who lean on AI to polish grammar. But with AI's tendency to fabricate facts and references, researchers are calling for clear guidelines rather than outright prohibition. The question isn't whether to use AI, but how to use it responsibly.

近日，有消费者发现SK海力士在淘宝天猫的官方旗舰店突然宣布关店，所有商品已下架，引发对已购内存保修问题的担忧。记者调查发现，各店铺独立运营，售后由店铺自行负责，部分商家提供五年质保。淘宝客服建议，若店铺关闭，可尝试通过订单详情申请售后，或联系平台客服，也可查询品牌官方保修。此次事件再次提醒消费者，网购存储产品时需关注品牌授权和售后责任。

字节跳动正式发布AI办公助手豆包，它不仅能写文档、做表格，还能在后台自动执行复杂任务。更关键的是，豆包深度整合了飞书，能利用企业内部数据，让AI真正懂业务。目前，豆包已应用于项目周报、客户需求分析等场景，并推出30天免费试用。

Goldman Sachs partner Chris Chapman has raised concerns that Wall Street's heavy reliance on AI might be eroding the analytical skills of young finance professionals. In a recent podcast interview, he warned of 'cognitive atrophy' as algorithms take over complex reasoning. Chapman compared this to how GPS has weakened our navigation skills, and he worries that the mentorship culture that trains junior bankers is at risk. His comments spark a crucial debate about balancing technological efficiency with human expertise.

WikiHow, the go-to site for step-by-step guides, has taken OpenAI to court. The lawsuit claims OpenAI scraped over 11,000 WikiHow articles to train ChatGPT, potentially violating thousands of copyrights. This case could redefine how AI uses online content, raising big questions about fair use and the future of content creation.

Moonshot AI announced that Kimi K2.5, its first trillion-parameter multimodal model, will retire at the end of this month. Launched in January, K2.5 was a versatile model supporting vision and text, with thinking and non-thinking modes. Its successor, Kimi K3, open-sourced in July, boasts 2.8 trillion parameters and a 1M token context window, continuing Kimi's record for largest open-source models.

Alibaba's DAMO Academy, with Shengjing Hospital, developed an AI model called DAMO LiON that can detect tiny liver tumors on CT scans. In a two-month trial, it found 15 overlooked malignancies, mostly around 1 cm. The AI acts as a safety net, helping doctors catch what they might miss, and the research is published in Nature Medicine.