Anthropic studied how Claude keeps knowledge across sessions without changing the model itself. Its tests suggest that short, selective notes work better than huge archives, and that having Claude review its notes later can fix errors.
Anthropic has released a list of 10 free AI courses for 2026, covering basics, agent systems, business, and education. The courses aim to help learners understand AI tools and build practical skills.
This short video is about how fast AI-written code is outpacing people's ability to check it. The creator explains that many developers say they generate AI code faster than they can verify it. He describes his own setup, where AI agents open code changes all day and nothing gets merged until a check the agent doesn't control passes.
OpenAI has launched a computer-use agent for healthcare that works through existing software to handle multi-step office tasks like prior authorization, referrals, and discharge planning. It is offered through ChatGPT for Healthcare, ChatGPT for Clinicians, and the OpenAI API.
Picsart explained how its AI Agents now talk with users by voice, using OpenAI's GPT-Live. The creative work keeps running in the background even after the call ends. Picsart said the feature arrived in September.
ElevenLabs is acquiring Kotoba, a Japanese voice AI lab, to speed up the use of AI agents and voice tools in Japan. Kotoba's team has deep voice research and knowledge of the Japanese market. The companies plan to build voice AI that fits Japanese language and social customs.
Thariq Shihipar, an independent developer, talks about how vibe coding, the practice of building software by simply describing what you want to an AI, can leave creators unsure how their own work functions. He covers the Bitter Lesson (the idea that general computing methods beat hand-built rules over time), along with Bash, React, and MCP. The talk also includes a live demo of readable specifications and custom scripting.
Anthropic has released a list of 10 free AI courses for 2026, covering basics, agent systems, business, and education. The courses aim to help learners understand AI tools and build practical skills.
Chip design companies are starting to offer AI agents as reusable tools that can explore design options and improve chip layouts. The article says trust, security, and pricing models are still unsettled, and adoption will take time.
Amazon Bedrock AgentCore runtime instances are now available in six more AWS regions. Runtime instances run AI agents on Amazon EC2 servers, and AgentCore handles setup, updates, scaling, and lifecycle management.
This short video explains where an AI agent runs once it leaves your laptop, using Google Cloud Run. It covers when to pick a service, a job, or a worker pool.
OpenAI has announced new ways to use ChatGPT for healthcare, including computer-use agents, improved clinical answers, and healthcare plugins that work across healthcare software.
Fabian Hedin, CTO of Lovable, explains at CODE Summit 2026 how the company's first AI agent framework nearly failed and what changed afterward. The talk covers durable agent runtimes, Git-style histories, and forkable computers for long-running coding agents.
ZTE, a Chinese telecom company, has put about 80 'digital employees' to work in its operations. The source material is a short excerpt about warehouse work at its Nanjing base.
Google has released Gemini Decision 1, a decision-making AI model. Decision models guide the judgments of AI agents, working differently from large language models (LLMs).
Grok Bot, an AI agent from SpaceX Superintelligence (SpaceXSI), can join Google Meet, Zoom, and Microsoft Teams calls to listen, take notes, and answer questions in meeting chat. After a call, it sends a summary and action items. The meeting feature works on Team, Enterprise, and Ultra plans.