Welcome back, knticNodes. Today, AI updates show how artificial intelligence is rapidly becoming part of both daily work and cutting-edge research. Google is embedding Gemini AI across Workspace apps to boost productivity, while NVIDIA’s new open model advances robotics and physical reasoning.
OpenAI launched Prism, a collaborative scientific writing platform, and DeepMind introduced AlphaGenome to explore gene regulation at scale. At the same time, AI is now helping build better AI, and Anthropic revealed new ethical principles to guide safer, more transparent systems.
by MIDJOURNEY
Google Expands Gemini Integration Across Workspace
Google has expanded the availability of Gemini across Google Workspace, integrating its AI capabilities directly into apps like Gmail, Docs, Sheets, Slides, and Meet through new Gemini Business and Gemini Enterprise plans. These features are designed to help teams draft content, summarize information, analyze data, and support everyday work tasks directly within Workspace. Gemini can assist with activities such as writing emails, creating documents, organizing information, and generating insights, helping teams improve productivity through AI-powered assistance embedded in their existing tools.
Source:
NVIDIA just released Cosmos Reason 2, a new open vision-language model designed for physical AI.
Cosmos Reason 2 focuses on stronger physical reasoning, spatial understanding, and long-context processing, supporting inputs of up to 256K tokens. It’s built to help robots and embodied AI systems better interpret visual scenes, reason about the physical world, and plan actions more effectively. NVIDIA positions the model as a foundation for robotics research, simulation workflows, and real-world decision-making, especially where understanding space, motion, and context really matters.
The model is open and available on Hugging Face, so you can explore it directly, run examples, and experiment with physical reasoning tasks without needing specialized hardware.
Source:
Back in 1943, a naval engineer named Richard James was trying to create springs that could help stabilize delicate equipment on ships rocking in the ocean. One day in his lab, he accidentally knocked one of his prototypes off a shelf.
But instead of crashing straight to the floor, the spring did something wild — it “walked.” It stepped down from a book to a stack of papers, then onto the table, and finally landed standing upright on the floor.
James went home excited and told his wife, “I think if I get the right kind of steel and the perfect tension, I can make this thing walk.”
And that’s how the Slinky was born. They first showed it in a department store in Philadelphia during Christmas of 1945 — and sold their first 400 in just 90 minutes
https://www.museumofplay.org/toys/slinky/OpenAI has released Prism, an AI-native workspace built for scientific writing and collaboration.
Prism is a LaTeX-native environment that embeds GPT-5.2 directly into the writing process, helping researchers and technical teams draft, revise, and collaborate on scientific documents in one place. The platform is designed to support tasks like structuring papers, refining language, working with equations, and iterating on technical content, all within a shared, real-time workspace.
Prism is available as a free tool and is aimed at making scientific and technical writing more fluid, especially for teams working on research papers, technical reports, or exploratory analysis where clarity and iteration matter.
Source:
Google DeepMind just unveiled AlphaGenome, an AI model that helps researchers understand how DNA controls gene regulation.
GBY GROKROK
The model can read up to one million DNA letters at once, making it easier to study how genetic mutations might affect biological processes linked to disease. Trained on human and mouse data, AlphaGenome is designed as a research tool to help scientists focus their lab work on the most relevant genetic variants.
It’s an early but promising step toward using AI to explore the genome at scale.
Source:
A new report from Georgetown’s CSET looks at how AI is starting to help build better AI.
The study explains how labs like OpenAI and Google DeepMind are using AI tools to automate parts of AI research, from testing models to speeding up experiments. It’s not about self-improving AI yet, but about making AI development more efficient.
If you’re tracking where advanced AI is headed, the full CSET report is a quick, useful read.
Source:
We hope you enjoyed this edition of kntic. Our focus is on applied AI, real-world impact, and how teams turn agents into results.
See you.
— kntic.