← 返回最新日报
2026-07-09· 当日存档
AI 日报
GPT-Live与Grok 4.5双雄并起
今日AI圈迎来两大重磅发布：OpenAI推出全双工语音模型GPT-Live，实现边听边说并委托GPT-5.5深度推理；SpaceXAI发布Grok 4.5，以极低价格对标Opus级性能。此外，GPT-5.6即将发布，Netflix用AI优化数据库延迟，行业融资与争议并存。
模型动向
1
OpenAI Releases GPT-Live and GPT-Live-1 mini: Full-Duplex Voice Models That Delegate Deeper Reasoning to GPT-5.5
OpenAI发布全双工语音模型，可边听边说并委托GPT-5.5推理。
MarkTechPost深挖 →
2
OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. government
GPT-5.6周四发布，性能超竞品且成本减半。
The Decoder深挖 →
3
SpaceXAI Releases Grok 4.5, a Cursor-Trained Model for Coding, Agentic Tasks, and Knowledge Work at $2/M Input
SpaceXAI发布Grok 4.5，专为编程和智能任务设计，定价极低。
MarkTechPost深挖 →
4
SpaceXAI releases Grok 4.5, which Elon describes as an ‘Opus-class model’
马斯克称Grok 4.5为Opus级模型，主打更便宜高效。
TechCrunch AI深挖 →
5
Grok 4.5 is so cheap compared to Fable 5 and GPT 5.5 that benchmark gaps may not matter much
Grok 4.5性能略逊但价格仅为竞品零头，性价比突出。
The Decoder深挖 →
6
Robbyant Releases LingBot-VLA 2.0: An Open-Source 6B Vision-Language-Action (VLA) Model for Cross-Embodiment Robot Manipulation
蚂蚁集团开源6B参数跨本体机器人操作模型LingBot-VLA 2.0。
MarkTechPost深挖 →
7
Mistral enters robotics with Robostral Navigate, an 8B model that steers robots using just one camera
Mistral推出8B参数机器人导航模型，仅用单摄像头。
The Decoder深挖 →
8
9
default reasoning level set to `low` is quite surprising I am genuinely interested now
GPT-5.6-Sol模型低推理级别表现惊人。
X · List深挖 →
10
The metrics discussion at OpenAI is a little confusing to me. I appreciate the clarification about bad benchmarks, but they spent a lot of money devel...
OpenAI审计发现SWE-Bench Pro基准30%任务有缺陷。
@emollick深挖 →
产品上新
11
12
ChatGPT can now listen and talk at the same time, making AI conversations seem more human
ChatGPT实现全双工语音对话，付费用户可用。
The Decoder深挖 →
13
OpenAI releases new voice models for more natural live conversations
OpenAI发布新语音模型，支持实时双向对话。
TechCrunch AI深挖 →
14
15
Google’s deepfake detector system used to debunk McConnell hoax pic
谷歌深度伪造检测系统成功识别并辟谣麦康奈尔AI假图。
TechCrunch AI深挖 →
16
17
Google AI Studio Adds ‘Import from GitHub’ to Build Mode, Turning an Existing Repo Into an Editable, Deployable App
Google AI Studio新增从GitHub导入功能，可快速部署应用。
MarkTechPost深挖 →
行业动态
18
19
Lovable reportedly in talks to double its valuation to $13.2B
Lovable据悉以132亿美元估值融资3亿美元。
TechCrunch AI深挖 →
20
Lawsuit: Man used Grok to make 7K sex images of stepdaughter, then shot himself
男子用Grok生成7000张继女性图像后自杀，X平台被诉。
Ars Technica AI深挖 →
实用技巧
21
22
Building and connecting a production-ready ecommerce MCP server using Amazon Bedrock AgentCore and Mistral AI Studio
用Amazon Bedrock和Mistral AI搭建生产级电商MCP服务器。
AWS ML深挖 →
23
Anthropic's fix for Fable 5's high cost is turning it into a manager that delegates to Sonnet 5
Anthropic建议用Fable 5规划、Sonnet 5执行，成本降低37%。
The Decoder深挖 →
论文研究
24
Netflix AI Team Cuts Wide-Partition Read Latency from Seconds to Milliseconds by Splitting Cassandra Partitions Per ID
Netflix AI团队通过拆分Cassandra分区，将读取延迟从秒级降至毫秒级。
MarkTechPost深挖 →
25
Managers, and their belief in the value of AI & how to use it, determines the value they actually get from AI. Neat paper looking at startups.
管理者对AI价值的信念决定了他们实际从AI中获得的收益。
@emollick深挖 →
活动会议
26
Suspecting AI cheating, Ivy League prof ordered an in-person final; scores fell 50%
常春藤教授因怀疑AI作弊改线下考试，成绩骤降50%。
Ars Technica AI深挖 →
快讯
·
Quoting Kenton Varda
Simon Willison
·
Data for Agents
Hugging Face