![[a0105d61-bdda-4c14-b489-a71d5a48cade.mp3]]

抖音生活服务上线AI解决方案提升商家获客与接待效率

简报：

• 2025年6月19日，抖音生活服务在武汉发布全新“AI即用”智能化解决方案，旨在帮助商家提升接待效率。
• 同步推出全新“私信获客”模式和产品功能，助推线索型商家扩大获客规模、提高客资质量。
• 过去一年“私信”业务用户量增长34%，对话量增长25%，目前90%以上线索商家已启用该功能，成为汽车、家居等行业主要获客渠道。

相关链接：

• 36氪快讯：抖音生活服务发布“AI即用”智能化解决方案
• AIBASE报道：抖音生活服务“AI即用”方案发布

Douyin Life Services Launches AI Solutions to Boost Merchant Customer Acquisition and Reception Efficiency 🚀

Brief:

• On June 19, 2025, Douyin Life Services unveiled its new "AI Instant-Use" intelligent solution in Wuhan, aiming to help merchants enhance their reception efficiency. 🤖
• Concurrently, a new "Private Message Customer Acquisition" model and product features were introduced to assist lead-based merchants in expanding their customer acquisition scale and improving lead quality. ✨
• Over the past year, the "Private Message" business saw a 34% increase in user volume and a 25% increase in conversation volume. Currently, over 90% of lead merchants have enabled this feature, making it a primary customer acquisition channel for industries such as automotive and home furnishing. 📈

Related Links:

• 36kr Newsflash: Douyin Life Services Releases "AI Instant-Use" Intelligent Solution
• AIBASE Report: Douyin Life Services "AI Instant-Use" Solution Launched

OpenAI或于2024年夏发布新一代AI模型GPT-5，引发行业高度关注

简报：

• OpenAI首席执行官Sam Altman近日表示，下一代大型语言模型GPT-5“可能将于今年夏季发布”，但未透露具体细节。
• 业界普遍预期GPT-5将在多模态、上下文理解与推理能力等方面实现突破，并需应对日益严格的合规要求。
• GPT-5的发布被视为AI产业技术和市场格局的重要拐点，其性能表现预计将影响行业领头羊地位和相关企业市值。
• Altman的表态推动资本市场关注，竞争对手包括谷歌、Anthropic等也在加速新一代大模型研发。
• GPT-5的实际上线时间尚未最终确定，OpenAI仍坚持产品发布前的保密策略。

相关链接：

• OpenAI CEO：大模型GPT-5或夏季发布，行业聚焦其技术突破与商业前景
• OpenAI CEO透露GPT-5或今夏上线，或重新定义人工智能边界

OpenAI Likely to Release Next-Gen AI Model GPT-5 in Summer 2024, Drawing High Industry Attention

Brief:

• OpenAI CEO Sam Altman recently stated that the next-generation large language model GPT-5 "might be released this summer," but did not disclose specific details. 📅
• The industry widely expects GPT-5 to achieve breakthroughs in multimodal capabilities, contextual understanding, and reasoning, while also needing to address increasingly stringent compliance requirements.
• The release of GPT-5 is seen as a crucial turning point for the AI industry's technological and market landscape. Its performance is expected to impact the leading position in the industry and the market capitalization of related companies. 🚀
• Altman's statement has drawn attention from capital markets, with competitors like Google and Anthropic also accelerating the development of their next-generation large models.
• The actual launch date for GPT-5 has not been finalized, and OpenAI continues to maintain a confidentiality strategy before product releases. 🤫

Related Links:

• OpenAI CEO: GPT-5 Large Model Expected This Summer, Industry Focuses on Its Technical Breakthroughs and Commercial Prospects
• OpenAI CEO Reveals GPT-5 Could Launch This Summer, Potentially Redefining AI Boundaries

谷歌AI推理Token一年增长50倍，领先全球同行

简报：

• 2024年至2025年4月间，谷歌人工智能推理的Token消耗量激增至480万亿，是微软和ChatGPT的5倍以上（后两者为100万亿）。
• 这一快速增长主要受到谷歌推出多款免费AI产品（特别是搜索引擎AI概述功能）的推动，谷歌AI用户基数为ChatGPT的5到6倍。
• 尽管AI推理成本上升，但其支出仍只占谷歌搜索收入的1%，公司继续加强AI研发和新模型训练。
• 谷歌计划推出更多创新AI应用，如通用助手、浏览器代理和视频生成模型，预示Token消耗量将进一步扩大。

相关链接：

• 谷歌AI推理Token激增50倍：数据与行业分析
• 谷歌AI消费量超微软OpenAI万亿规模
• IT之家报道：谷歌Search Live语音搜索创新

Google AI Inference Token Usage Soars 50-Fold in a Year, Leading Global Peers

Brief:

• Between 2024 and April 2025, Google's AI inference token consumption surged to 480 trillion, more than 5 times that of Microsoft and ChatGPT (which stand at 100 trillion each). 🚀
• This rapid growth is primarily driven by Google's launch of multiple free AI products, especially the AI Overviews feature in its search engine. Google's AI user base is 5 to 6 times larger than ChatGPT's.
• Despite the rise in AI inference costs, spending on it still accounts for only 1% of Google's search revenue. The company continues to strengthen AI R&D and new model training. 💪
• Google plans to introduce more innovative AI applications, such as a universal assistant, browser agent, and video generation models, signaling a further expansion in token consumption. ✨

Related Links:

• Google AI Inference Tokens Soar 50x: Data and Industry Analysis
• Google's AI Consumption Exceeds Microsoft OpenAI's Trillions
• IT Home Report: Google Search Live Voice Search Innovation

Midjourney发布首款图像转视频AI模型V1，月费10美元、最长支持21秒生成

简报：

• Midjourney正式上线首个AI视频生成模型V1，主打图片转视频（Image-to-Video）能力，用户可将静态图片转换为最长21秒的视频，每月最低10美元即可体验。
• V1模型支持两种不同风格的运动模式，允许自定义视频时长，每次生成可产出四段不同风格片段，画质为480p/24帧。
• 当前版本暂不支持超分辨率和音频输出，继承自其V6.1图像模型的风格和一致性，面向普通和专业创作者提供不同订阅方案。
• V1发布后受到了用户的积极评价，因操作简便、视觉细腻而受到好评，尽管高动态场景仍有轻微闪烁技术问题需优化。
• 同时，Midjourney正面临迪士尼和环球影业的版权诉讼，因其AI训练涉及受版权保护的角色素材。

相关链接：

• Midjourney视频生成模型V1正式发布，最长21秒，月费10美元
• Midjourney推出首个视频AI模型，打造完整创意闭环
• Midjourney新模型引发市场热议，被指涉版权风险

Midjourney Releases First Image-to-Video AI Model V1: $10/Month, Up to 21-Second Generations

Briefing:

• Midjourney officially launched its first AI video generation model V1, focusing on Image-to-Video capabilities. Users can convert static images into videos up to 21 seconds long, available for as low as $10 per month. 🚀
• The V1 model supports two distinct motion styles, allows for custom video durations, and generates four different stylistic clips per creation, with a resolution of 480p/24fps.
• The current version does not yet support super-resolution or audio output. It inherits the style and consistency from its V6.1 image model and offers different subscription plans for both general and professional creators.
• Following its release, V1 has received positive user feedback, praised for its ease of use and visual sophistication, though slight flickering in high-dynamic scenes still requires optimization. ✨
• Concurrently, Midjourney is facing copyright lawsuits from Disney and Universal Studios, as its AI training involves copyrighted character materials. ⚖️

Related Links:

• Midjourney Video Generation Model V1 Officially Released: Up to 21 Seconds, $10 Monthly Fee
• Midjourney Launches First Video AI Model, Creating a Complete Creative Loop
• Midjourney's New Model Sparks Market Buzz, Accused of Copyright Risk

合合信息推出首个AI Agent云资源智能管理终端Chaterm

简报：

• 合合信息在2025亚马逊云科技中国峰会上发布业内首个AI Agent跨平台云资源智能管理终端Chaterm。
• Chaterm通过对话式终端管理工具，支持云计算从业者用自然语言实现云资源的智能化、规模化管理，并可自动完成开发与运维等任务。
• 该产品的核心代码现已完全开源。

相关链接：

• 36氪新闻报道
• aibase新闻1
• aibase新闻2

Hehe Information Unveils Chaterm: The First AI Agent for Smart Cloud Resource Management

Briefing:

• Hehe Information announced Chaterm, the industry's first AI Agent cross-platform cloud resource intelligent management terminal, at the 2025 AWS China Summit. 🚀
• Through its conversational terminal management tool, Chaterm empowers cloud computing practitioners to achieve intelligent, scalable management of cloud resources using natural language, and can automatically complete development and operations (DevOps) tasks. 💬
• The product's core code is now fully open-sourced. 💻

Related Links:

• 36Kr News Report
• aibase News 1
• aibase News 2