NotebookLM Video Overviews will get new styles, including Anime

Google is expanding NotebookLM's video overview feature with new visual styles and a format selector, signaling broader customization options. These updates align with upcoming AI model improvements and a larger product rollout strategy.

🗞 #notebooklm

Google is working on Formats and visual Styles support for Video Overviews on NotebookLM.

Users will be able to generate Video Overviews in Anime, Heritage, Watercolour and other styles.

Jules SWE Agent now has an API! Users can now generate an API key in order to integrate Jules into their workflows and apps.

Google is preparing Veo 3 "Frames to Video" for release on Flow, as Veo 2 is about to get decommissioned.

This is the best example I've ever seen 👀
- thx to @wrg133

ICYMI: Perplexity launches Comet AI browser worldwide for free

Perplexity’s AI browser Comet is now globally available, removing its waitlist. Designed for desktop research and productivity, it includes assistants for tasks, background processing, and voice input, with mobile apps in preview and subnoscriptions offered.

🗞 #perplexity

BREAKING 🚨: OpenAI acquired Roi, a personalised finance AI assistant app.

"The products we use every day won’t remain static, predetermined experiences. They’ll become adaptive, deeply personal companions that understand us, learn from us, and evolve with us."

OpenAI added a new Codex Alpha opt-in option in Codex settings.

"Get early access to new Codex models and features."

Early Access for free!!! 👀

Google tests fully autonomous plan in Jules agent

Google is testing a “Fully Autonomous Plan” mode in its engineering agent Jules, allowing it to manage entire development tasks without human input.

🗞 #jules

OpenAI to add granular rights controls to Sora video platform

OpenAI is updating Sora 2 to provide rightsholders with detailed controls over character generation, allowing them to specify conditions for use of likenesses. These changes aim to address IP concerns and are pending broader release.

🗞 #sora

Google A/B tests Gemini 3 on AI Studio ahead of October 9 release

Google is set to launch Gemini 3 on October 9, showcasing major improvements in SVG generation, UI features like “My Stuff,” and deeper tool integration. Early benchmarks show it outperforming rivals in technical and coding tasks.

🗞 #gemini

Exclusive: Opera launches its Agentic AI browser Neon to early testers

Opera's Neon browser launches as a $59, nine-month subnoscription targeting early adopters, offering AI-powered features like multi-mode prompts, automation tools, and creative workflows, with early access and direct developer feedback.

🗞 #opera

BREAKING 🚨: Exclusive, Early, $59-gated, first look at Neon agentic AI browser from Opera.

It comes with 4 different modes:
- neon - traditional browsing
- chat - AI chat
- do - Agentic browsing and tasks
- make - generating applications, images, and canvases

Alpha mode 👀

OpenAI prepares to release Agent Builder during DevDay on October 6

OpenAI’s upcoming Agent Builder offers a visual tool for building AI workflows, targeting developers and businesses using OpenAI APIs. It supports modular logic, prebuilt templates, and integrates tightly with OpenAI’s platform and security.

🗞 #chatgpt

BREAKING 🚨: OpenAI is about to add a model selector to the Images tab on OpenAI Platform.

New image model tomorrow? 👀

And loads of guardrails 👀

PII, Moderation, Jailbreak, Hallucination, Prompt Injection and more.

Agen Builder preview 👀

OpenAI may release new Image model and Guardrails soon

OpenAI is testing new platform features, including a model selector for image generation and a Guardrails section for API safety. These updates aim to support cost efficiency, regulatory compliance, and broader use across creative and enterprise needs.

🗞 #chatgpt

One interesting feature in Agent Builder is ChatKit and .widget support.

It also refers to an application called Widget Studio where users will be able to develop UI widgets for their agents.

A big differentiation point 👀