1440: Your Weekly Business Cheat Sheet

Expand your business and finance knowledge with 1440. Get clear, conversational breakdowns of the key concepts in business and finance—no paywalls, no spin. Every Thursday, 1440 delivers deep dives, interactive charts, and rapid market rundowns trusted by 100k+ professionals.

Sign Up Today

Today:

• Google’s Kingfall Leak Stuns AI

• Reddit Sues Anthropic Over Scraping

• ChatGPT Adds Drive, Meeting Tools

• Mistral Code Rivals GitHub Copilot

• Claude Pro Unlocks Premium Tools

Google's Secret "KINGFALL" Model Leaked... plus other AI News

A confidential Google AI model called Kingfall briefly appeared and impressed users with its deep thinking mode and high token capacity. It may rival OpenAI’s upcoming o3 Pro. 

Meanwhile, a movie about Sam Altman's 2023 OpenAI firing is underway, and major bonuses were offered to retain researchers. Claude AI launched a blog, OpenAI added new productivity features, and Anthropic limited Claude access on Windsurf—possibly signaling growing platform competition.

WATCH THE VIDEO ON YOUTUBE

Reddit Sues Anthropic Over Scraping

Reddit has sued Anthropic for taking, without payment, huge amounts of user posts to train its chatbot. The lawsuit says Anthropic tapped Reddit’s servers over 100,000 times, ignored offers to license the data and profited unfairly. Reddit, now selling its conversation data to Google and OpenAI, argues that data use must be paid for and controlled to protect users’ privacy and the site’s value. Case could reshape rules.

Why it matters

• Who pays for data? The suit could set the first clear rule that AI labs must buy paid permission (“a license”) before using public conversations to train models.
  

• Privacy spotlight. Reddit is demanding that users keep the right to delete posts and prevent spam, pushing AI builders to respect personal content.
  

• Cost of training. If courts side with Reddit, top-quality data will get pricier, forcing labs to strike new deals or rethink how they teach their systems.

ChatGPT Adds Drive, Meeting Tools

OpenAI is beefing up ChatGPT for business users. The tool can now search files stored in Google Drive, Dropbox, Box, OneDrive, and SharePoint, following company permission rules. It also records and transcribes meetings, creates time-stamped notes, and turns tasks into Canvas docs. New “deep research” connectors tap HubSpot, Linear and more, while an open Model Context Protocol lets firms wire up apps. The goal: make ChatGPT the center of work.

Why it matters

• More connectors let ChatGPT work directly on a company’s own documents, bringing AI into daily workflows instead of keeping it separate.
  

• Automatic meeting recording and note-taking show how AI now handles routine office chores, freeing people for creative thinking.
  

• The open protocol for linking extra apps hints at a future where one AI can talk to every business tool, making work systems more flexible.

Mistral Code Rivals GitHub Copilot

Mistral AI has launched “Mistral Code,” a coding helper for big companies. It runs on company hardware (on-premise) so secret code never leaves. Firms can fine-tune the helper on their projects, boosting accuracy for special workflows. The tool bundles models, editor plugins, and 24/7 support under one contract. Early users in banking and rail show it meets strict privacy rules, challenging GitHub Copilot and other U.S. rivals in fast-growing market.

Why it matters

• Keep your data home. On-premise setup answers privacy worries and supports “data sovereignty” — full control over your own data.
  

• Tailored brains. Companies can train the assistant on their private code, proving that future AI must adapt to each firm instead of offering one flat solution.
  

• Healthy rivalry. A strong European alternative to U.S. tools, shaped by strict EU rules, pushes innovation and shows smaller, locally run models can compete.

🧠RESEARCH

MiMo-VL Technical Report

Xiaomi’s MiMo-VL-7B models beat larger AI systems in understanding images and text together. Using advanced training and reinforcement learning, they lead in over 35 tests, including tough reasoning and interface tasks. The team also shares tools and data to help others build and test similar models.

A Controllable Examination for Long-Context Language Models

LongBioBench is a new test that uses fake biographies to fairly judge how well language models handle long texts. It improves over older tests by being more realistic, easier to control, and better at spotting gaps in understanding, reasoning, and trust. Most models still struggle as context length grows.

SuperWriter: Reflection-Driven Long-Form Generation with Large Language Models

SuperWriter-Agent helps AI write better long texts by planning and refining like a human writer. Using a 7B model and smart training methods, it beats larger models in quality and coherence. Structured thinking and step-by-step improvements make its writing more logical, consistent, and professional.

🛠️TOP TOOLS

GPTKit  - AI-powered text detection tool designed to distinguish between human-written and AI-generated content with high accuracy.

ImageToCartoon - AI-powered online tool that transforms photos into cartoon-style images quickly and easily.

Wonder Dynamics - AI-powered visual effects company that revolutionizes the film and entertainment industry.

Watermark Remover IO - AI-powered online tool designed to efficiently remove watermarks, logos, text, and other unwanted elements from images while preserving the original quality.

Typeframes - AI-powered text-to-video creation tool designed to simplify the process of producing engaging video content for platforms like YouTube, Instagram, and TikTok.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Anthropic’s Claude Pro now includes three advanced tools—coding via command line, app integrations like Jira and Asana, and deep research from web data. These features were previously limited to higher-tier plans, now free for Pro users.

• OpenAI reached 3 million enterprise users and launched workplace tools that rival Microsoft, including data connectors, meeting transcription, and advanced coding agents. These moves cement its leadership in business-focused AI services.

• Anthropic launched a blog written by its AI, Claude, with human editors refining the content. The project showcases how AI can support, not replace, expert writers—amid rising industry use of generative tools and ongoing concerns about AI accuracy.

• Amazon will invest $10 billion in North Carolina to expand AWS data centers, create 500 skilled jobs, and support thousands more. The plan includes AI infrastructure, education programs, and a $150,000 community fund for local development.

• Brookfield will invest nearly $10 billion in an AI-focused data center in Strängnäs, Sweden, creating 1,000 permanent and 2,000 temporary jobs. The project highlights Europe’s AI infrastructure boom driven by energy reliability and data sovereignty laws.

• Amazon has created a new agentic AI group inside Lab126 to build intelligent systems that let robots understand and act on human language. The effort focuses on robotics and "physical AI," signaling a deeper move into advanced automation.