Gemini Agent is presented as a single, enterprise-ready AI assistant that handles questions, knowledge work, content and code generation, and autonomous tasks from one prompt box and one API. It runs inside Workspace apps (Gmail, Drive, Docs, Slides, Sheets, Chat, Calendar) and across devices, maintaining persistent execution, memories, and a personalization graph so work can continue after users disconnect. Architecturally it supports multi-agent orchestration (temporary sub-agents and persistent coworker agents with their own identities), unified skill and tools registries, and deep context from connected systems (BigQuery, Snowflake, Salesforce, Git, Jira, Office, Slack, etc.). Model choice is flexible: jobs are routed across Gemini-family models and third‑party models (Anthropic Claude today) to balance accuracy and cost, with Smart Routing and real-time spend caps.
Enterprise specifics emphasize secure governance and integration: identity and policy management, authorization, secure sandboxing, network gateways, and tool-level access controls. Skills (reusable modular workflows) and four memory types (session, semantic, procedural, episodic) let agents learn procedures and company knowledge. Customer outcomes cited include BNP Paribas, Bradesco (document review cut from 1 hour to 5 minutes, 60% fewer inconsistencies, >10% financial efficiencies), Ulta Beauty (3x digital conversion), Orange Spain (>1,000 custom agents), Sompo (10,000 agents), Shopify and PayPal multi-model routing, and projected modeling speedups up to 75% for utilities - demonstrating productivity, automation scale, and cost/governance focus.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Four coding agents were given a $100 budget to build a functional PDF editor, but their implementations had noticeable bugs and shortcomings. The experiment highlights that current AI agents struggle with goal pursuit and user interaction, leading to subpar software quality. (blog.nielstron.de)
Lithos-ai open-sourced lithos-metal, enabling Qwen3.8-27B to run at over 200 tokens per second on an Apple M5 Max laptop. This allows ultra-fast inference for coding agents with a single command. (twitter.com)
Temporal developed pi-temporal to enable the Pi coding agent to recover from machine failures during tasks by leveraging a fleet of workers and durable execution. This setup allows seamless task continuation across machine failures without repeating interrupted commands. (temporal.io)
![OpenAI cannot make AI safe on its own [pdf]](https://hntoday.pixelshot.io/screenshots/20261008/0217d5d1410143c5_tn.jpeg)
OpenAI has acknowledged that it cannot ensure the safety of artificial intelligence systems on its own. The document discusses the challenges and limitations faced by OpenAI in making AI safe without external collaboration or oversight. (mikitabalesni.com)
Yoshua Bengio urges AI researchers at frontier companies to prioritize safety and consider leaving if safety measures are insufficient. He warns that current development races risk pushing humanity toward catastrophic outcomes and calls for slowing AI progress to prevent loss of control. (transformernews.ai)
A new open vision-language model has been developed for various medical applications, enhancing the integration of visual and textual data in healthcare. The model aims to improve diagnostic accuracy and support medical decision-making through advanced AI techniques. (nature.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.