📊 Claude lands inside Microsoft Excel for Pro users

Anthropic has встроила Claude directly into Microsoft Excel. The integration lets users work with data through plain tasks instead of writing formulas.

You can ask things like grouping customers by LTV and purchase frequency. Claude analyzes the table and returns results inside Excel. It also generates formulas for a given task, explains how they work, checks existing formulas for errors, builds pivot tables, compares periods, and flags trends or outliers.

The shift moves Excel from manual logic to intent based analysis. For Pro users, spreadsheets become a conversational interface rather than a syntax problem.

📊 @tech

📄 OpenAI launches Prism for writing research papers

OpenAI has released Prism, a collaborative LaTeX editor built on technology from its acquired startup Crixet. The tool embeds ChatGPT directly into the writing workflow.

Prism can check citations, turn sketches into LaTeX formulas or diagrams, and assist with drafting inside the editor. It is positioned as a working environment rather than a separate chat tool.

The product is free to use. Subscribers to ChatGPT are expected to get additional features later.

📊 @tech

🤖 Elon Musk says Tesla will start selling humanoid robots by end of next year

Elon Musk said Tesla plans to begin robot sales by the end of next year and stressed the timeline is firm. According to him, the humanoids already handle simple tasks at factories.

The robots are positioned for basic work and household routines like cleaning. Musk described their scope broadly, saying users will be able to ask them to do almost any task.

A test production line is already running at Tesla’s Fremont factory. Pricing is still not final, though earlier estimates were around $20,000 per unit.

📊 @tech

🌐 Chrome turns into an AI agent with Gemini integration

Google rolled out a major update that deeply integrates Gemini into Google Chrome. The browser now includes a persistent Gemini side panel that stays open across tabs and works alongside whatever page is active.

Users can compare products across sites, summarize reviews, or check availability while browsing.

Chrome also adds in-browser image transformation, connected access to Gmail, Calendar, YouTube, Maps, and shopping tools, plus an auto browse mode that handles multi step tasks like price research, form filling from PDFs, or adding items to a cart with budget limits.

The update shifts Chrome from passive browsing to task execution. For now, advanced features are limited to AI Pro and Ultra subscribers in the US.

📊 @tech

🎬 xAI launches Grok Imagine API for video generation

xAI has launched the Grok Imagine API. It is a video and audio generative model. The release was announced on xAI’s official X account.

Grok Imagine can animate images, generate scenes from text prompts, and refine longer cinematic sequences. The model is offered through xAI’s video API.

xAI positions this as a high performance API built for video workloads. The focus is on turning text or images into finished video output.

The launch puts xAI directly into the video generation model category.

https://x.ai/news/grok-imagine-api

📊 @tech

✔️ AI takes over homework grading in Chinese schools

Chinese schools are increasingly using AI to check homework. Systems scan handwritten notebooks, find mistakes, and assign grades automatically.

The technology is already used in hundreds of schools in Shanghai, Beijing, and other large cities. For essays and creative tasks, teachers often adjust results because algorithms still miss nuance.

According to developers, these systems cut teacher workload by almost half. Time shifts from routine grading to direct work with students.

📊 @tech

🔭 Vaonis launches a $99,000 telescope controlled from a smartphone

Vaonis has unveiled Hyperia, a smart observatory priced at $99,000. The system was originally built for the Palais de la Découverte in Paris before moving into a commercial release.

Hyperia uses Canon optics with a 150 mm refractor and a 1050 mm focal length. Inside is a 45 MP full-frame Canon sensor, comparable to EOS R5 II class cameras. A 17-lens system reduces reflections and supports long exposure imaging.

The structure stands about 1.8 m tall and weighs up to 75 kg. It can stay outdoors for up to 5 years without a dome or servicing. The system compensates for Earth’s rotation by rotating the sensor, enabling exposures of up to 30 minutes.

Everything runs through a mobile app. Users select targets, schedule observations across nights, process images, and stream live output to classrooms or planetariums.

Preorders opened in January. Deliveries are planned for 2027.

📊 @tech

🚀 AI plans a Mars rover route for the first time

On December 8, a Mars rover followed a route generated by AI for the first time in history. The test came from a collaboration between Anthropic and NASA.

Claude was loaded with years of rover context. This included telemetry, command formats, safety limits, orbital images, terrain models, and hazard data. The agent planned the path iteratively in 10 m segments.

The final route was reviewed by humans and tested in simulation. NASA says only minor adjustments were needed. Rover teams estimate AI planning cuts route planning time by about 2x and delivers more consistent results.

📊 @tech