Looking for unbiased, fact-based news? Join 1440 today.

Join over 4 million Americans who start their day with 1440 – your daily digest for unbiased, fact-centric news. From politics to sports, we cover it all by analyzing over 100 sources. Our concise, 5-minute read lands in your inbox each morning at no cost. Experience news without the noise; let 1440 help you make up your own mind. Sign up now and invite your friends and family to be part of the informed.

Subscribe to 1440 today.

Today:

• AI Will Soon Outthink Humans

• DeepMind Unveils AlphaGenome DNA Model

• Claude Adds Interactive Artifact Workshop

• Google Launches Free Gemini CLI

• Meta Wins AI Training Lawsuit

• Google Launches AI-First Colab

Dr Mike Israetel: We ALREADY Know How to Build ASI... Human Death Only Has DECADES Left

Dr. Mike Israel argues AI will outthink humans within years and quickly self-improve. Once models prompt themselves, gain memory, and rewrite their own “brain,” they reach ASI—artificial super intelligence. 

He expects robots, gene fixes, and mind uploads to end aging and routine work, while humans fuse with machines. He believes entropy—natural decay—is the real foe, so future AI will likely help humans thrive, not wipe us out, in his view.

WATCH THE VIDEO ON YOUTUBE

DeepMind Unveils AlphaGenome DNA Model

AlphaGenome is an AI model from Google DeepMind that reads up to one million DNA “letters” and predicts how any single change might affect gene activity. It covers both coding and non-coding regions, giving letter-by-letter insight into where genes start, stop, splice, and how much RNA they produce. This helps pinpoint disease-causing mutations, design custom DNA, and speed basic genome research. All released through a research API.

Why this matters

1. Pushes model limits — shows transformers can handle million-token biological sequences, guiding larger-context AI designs in other fields.
   

2. Bridges AI and medicine — offers a template for turning deep-learning insights directly into disease target discovery and drug design.
   

3. Open research catalyst — free API lets labs fine-tune or extend the model, encouraging community-driven innovation and benchmarking.

Claude Adds Interactive Artifact Workshop

Anthropic’s newest Claude update gives every user a workshop for “artifacts”—saved projects such as flashcards or mini-games—and now lets you drop Claude’s smarts directly inside them. Describe an idea and Claude builds a shareable, interactive app others can tweak. A sidebar hub displays, organizes, and remixes community creations, turning chat into making for learning, work, and play. No coding required; free, Pro, and Max users can test the beta today.

Why this matters

• Democratizes AI creation—spreads the power to build apps to anyone who can type, widening access and skill-building.
  

• Turns chat into code—shows a path where natural language becomes the main way to program, pointing to more human-friendly AI tools.
  

• Accelerates open innovation—a shared library of user-made apps sparks faster ideas, feedback, and model improvements across the community.

Google Launches Free Gemini CLI

Google’s free Gemini CLI puts its Gemini 2.5 Pro model inside the command line, developers’ daily workspace. Log in with a personal Google account and ask the model up to 1,000 times a day for code fixes, research, or task automation via chat. The tool is open-source—code anyone can inspect or modify—and extensible, links with Gemini Code Assist in VS Code, and supports commands, scripts, and web search grounding features.

Why this matters

• Open code raises trust — The tool is open-source (its code is public), so anyone can check or improve it, setting a clear example of transparent AI.
  

• Chat-style coding — Asking in plain English to create or fix programs hints at a future where talking replaces typing code, expanding who can build software.
  

• Roomy, free playground — The model’s million-token window (how much text it can read at once) and 1,000 daily calls let tinkerers test big ideas without paying.

🧠RESEARCH

AnimaX: Animating the Inanimate in 3D with Joint Video-Pose Diffusion Models

AnimaX is a new system that turns text and video into 3D character animations. It works with any type of skeleton and doesn’t need slow optimization. By blending video motion with pose control, it creates smooth, realistic movements. Trained on 160,000 examples, it outperforms other animation tools.

Matrix-Game: Interactive World Foundation Model

Matrix-Game is a powerful AI that creates high-quality Minecraft videos where users control actions and camera movement. Trained on 3,700 hours of gameplay, it understands both visuals and player input. It outperforms previous models in realism, precision, and rule-following, and will be released publicly to support future research.

Why Do Open-Source LLMs Struggle with Data Analysis? A Systematic Empirical Study

This study explains why open-source language models struggle with data analysis. The main issue is weak strategic planning, not just poor coding. Better interaction design and cleaner data make a big difference. Using these insights, the authors created improved training methods that boost the models’ ability to reason through complex analysis tasks.

🛠️TOP TOOLS

LogoAI - Creating professional logos, cohesive visual identities, and automating brand promotion.

BoredHumans - Offers over 100 free AI tools, designed to cater to a wide range of creative, entertainment, and productivity needs.

Duplichecker - Designed to ensure the originality and integrity of written content.

Mnml AI - AI-powered design assistant specifically designed for architects, interior designers, and related professionals.

Getimg - AI-driven image generation and editing tool designed to create high-quality visuals from text descriptions.

📲SOCIAL MEDIA

🗞️MORE NEWS

• A judge ruled Meta didn’t break the law by training AI on 13 authors’ books, citing no clear financial harm. But the decision applies only to this case and leaves room for future lawsuits.

• Google has launched an AI-first version of Colab for all users. It acts like a smart coding partner—cleaning data, fixing bugs, building charts, and explaining code—making machine learning and coding much faster and easier.

• WhatsApp now lets users generate AI-made summaries of missed messages using Meta’s technology. The optional feature, rolling out in the US, emphasizes privacy but raises questions about accuracy and growing AI integration in chat apps.

• U.S. lawmakers introduced a bill to ban Chinese AI tools, like DeepSeek, from use in federal agencies. Citing national security risks, the bill would create a list of foreign-developed AI and block their deployment.

• After Business Insider exposed security gaps, Scale AI quickly restricted access to internal documents involving major tech clients like Meta. The move highlights growing concerns over data protection in high-stakes AI partnerships.

• OpenAI warns that China’s Zhipu AI is rapidly expanding by securing government deals in Asia, Africa, and the Middle East. Backed by $1.4B in state funding, it reflects Beijing’s push to lead global AI development.

• Khan Academy CEO Sal Khan says AI will soon act like five top teaching assistants in every classroom—helping teachers personalize lessons, answer questions, and boost learning without replacing human educators.

• OpenAI and Microsoft are clashing over what qualifies as “artificial general intelligence” (AGI). Their deal gives OpenAI more control if AGI is reached—something Microsoft wants to avoid, sparking growing tension in their partnership.