Weekly AI #11

Welcome to our eleventh newsletter. The Stargate Project, a $500 billion AI infrastructure initiative backed by OpenAI, SoftBank, Oracle, and MGX, launches this week, aiming to build a network of data centers and solidify U.S. leadership in AI innovation.

Top 5 Tools

JoggAI - Easily create stunning videos with engaging AI Avatars

Trae - Trae is an adaptive AI IDE that transforms how you work, collaborating with you to run faster

ClipZap - AI workflow automation tools for your personalized videos

Dewstack AI - Documentation made easy with AI

Memex - Describe what you want to build in text, and Memex builds it iteratively with your feedback

Top News

DeepSeek-R1 Revolutionizes Open-Source AI - DeepSeek's new open-source AI model, DeepSeek-R1, is shaking up the AI landscape by offering competitive performance against proprietary models like OpenAI's o1. With its free-to-use MIT license and capabilities in reasoning tasks, it's fostering a decentralized, collaborative AI future.

Tech Giants Launch $500 Billion Stargate Project to Transform U.S. AI Infrastructure - A powerful coalition of tech giants, including OpenAI and Microsoft, is committing $500 billion to revolutionize U.S. AI infrastructure. With immediate investments and expanded data centers, they aim to create numerous jobs, bolster economic growth, and strengthen America's global technological position.

OpenAI Launches Operator, an AI Agent to Automate Web Tasks - OpenAI's Operator, an AI agent currently in research preview for U.S. ChatGPT Pro users, independently manages web tasks like booking travel or ordering groceries. By integrating advanced reasoning with website interaction, it promises to enhance digital experiences despite some early-stage limitations.

Shorts

Mon, Jan 20
- DeepSeek R1 distilled models now available in LM Studio
- LM Studio plans to add DeepSeek R1 support soon
- DeepSeek launches open-source DeepSeek-R1, challenging OpenAI’s dominance
- Sam Altman clarifies OpenAI has not developed AGI, despite speculation
- Liquid AI launches LFM-7B, a top multilingual language model

Tue, Jan 21
- Perplexity AI launches Sonar, an affordable search API
- Microsoft tests AI-based search engine in Windows 11
- Dario Amodei predicts human-level AI systems within 2-3 years
- Kling AI introduces Elements feature for creative experimentation
- ElevenLabs deploys Conversational AI agent for documentation support
- Hailuo AI introduces four advanced AI models for text, video, and audio
- Trump administration revokes AI regulations, contrasting EU’s AI Act
- Tencent launches Hunyuan3D-2.0, advancing 3D modeling with AI
- Midjourney updates moodboards with multi-code and sref mixing
- Kanye West seeks AI team members for new projects

Wed, Jan 22
- Anthropic CEO notes DeepSeek uses 50,000 H100s, while U.S. plans million-chip clusters by 2026
- Fal AI launches AI Video Starter Kit with industry partners
- Spline launches Spell, a model for 3D world generation
- OpenAI’s Stargate Project under scrutiny as Elon Musk questions $500 billion funding
- DeepSeek, a side project of a quant firm, uses GPUs for trading, mining, and AI development
- Microsoft CEO confirms $80 billion commitment to Project Stargate
- Trump announces Stargate Project, a massive AI infrastructure project
- Freepik AI Suite introduces MiniMax Reference mode for consistent AI videos
- Adobe Premiere Pro introduces AI-powered Media Intelligence for video search
- Google invests $1 billion more in Anthropic, totaling $3 billion
- Google AI Studio updates Gemini 2.0 Flash Thinking with new features
- Stargate Project focuses on AI-driven cancer detection and mRNA vaccines
- OpenAI announces Stargate Project with $500 billion investment for AI infrastructure

Thu, Jan 23
- Google DeepMind CEO predicts AGI with robust cognitive abilities in 3-5 years
- GitHub Copilot for Visual Studio now free with Claude 3.5 and GPT-4o access
- Pika 2.1 launching soon with 1080p and lifelike characters
- OpenAI launches Operator AI agent for web task automation
- Anthropic introduces Citations API for Claude’s source-based answers
- OpenAI expands free ChatGPT features with o3-mini model
- Perplexity Assistant launches for Android, transitioning to integrated assistant
- Freepik adds Google-powered Imagen 3 creations to AI Suite
- Elon Musk announces Grok 3, sparking excitement for its advanced features
- Krea AI introduces real-time custom AI model training
- ChatGPT outage sparks global frustration and humor over AI dependency
- Google DeepMind’s AlphaFold AI targets cancer, heart disease, and neurodegenerative treatments

Fri, Jan 24
- ChatGPT Canvas updates with o1 integration and React/HTML rendering
- Midjourney updates web describe feature for image analysis
- Kling AI upgrades KOLORS with Image Reference and precise control
- Qwen Chat v0.2 launches with web search, video, and image generation
- Meta plans $65 billion AI investment, including Llama 4 and massive data center
- Kling v1.6 video models now available on Replicate
- HeyGen introduces motion control for dynamic AI avatar videos

Sat, Jan 25
- Kimi.ai launches Kimi 1.5 on web with free unlimited usage and advanced features
- DeepSeek-R1 challenges global AI market with cost-effective open-source model

Sun, Jan 26
- LM Studio supports Qwen2.5-1M with 1M-token context for local use
- Qwen2.5-1M launches with 1M-token context and faster inference
- Hailuo AI offers early access to T2V 01-Director model for select users
- xAI develops voice mode for Grok app with ElevenLabs integration
- xAI enhances Grok with sports, finance, and location widgets