- H-FARM AI's Newsletter
- Posts
- xAI makes Grok 4 free for all users worldwide
xAI makes Grok 4 free for all users worldwide
PLUS: Google & NASA test AI for space medical care , Alibaba Qwen3 hits 1M context windows, US govt secures 15% cut from chip sales & AI companies poach Wall Street talent.

In today’s agenda: |
|
MAIN AI UPDATES / 11th August 2025
🤖 xAI makes Grok 4 free for all users worldwide 🤖
Xai introduces automatic routing system with expanded usage
xAI has launched Grok 4 as a free service for all users globally, marking a significant shift in AI accessibility. The platform features Auto mode that intelligently routes complex queries to Grok 4, while Expert mode allows users to manually select the advanced model.
For a limited time, xAI is rolling out generous usage limits to encourage exploration of Grok 4's full capabilities.
The move comes as xAI prepares its V7 foundation model, which finished pre-training last week and represents the first natively multimodal system that processes video and audio bitstreams directly without transcription. This breakthrough enables understanding of tonal nuances and mood that traditional text-based systems miss.
🚀Google and NASA test AI for medical care in space 🚀
Crew Medical Officer Digital Assistant could support Mars missions
Google and NASA have collaborated on an innovative proof-of-concept for an automated Clinical Decision Support System called the "Crew Medical Officer Digital Assistant" (CMO-DA). The AI system is designed to assist astronauts with medical help during extended space missions when crews are not in direct contact with Earth-based medical experts.
Trained on spaceflight literature, the system uses cutting-edge natural language processing and machine learning to provide real-time analyses of crew health and performance. Initial trials using the Objective Structured Clinical Examination framework showed promise for reliable diagnoses based on reported symptoms. The tool could benefit both space exploration and remote medical care on Earth.
🧠 Alibaba's Qwen3 models achieve 1M context windows 🧠
It can processes 1,500 pages simultaneously with 3x speed
Alibaba's Qwen3 AI models now support context windows of up to 1 million tokens, equivalent to analyzing approximately 1,200-1,500 pages of text simultaneously. The models are open-source and deliver 3x faster processing speeds than previous versions, putting pressure on rivals with smaller context windows.
The Qwen3-Coder variant enables developers to load entire codebases into the model's working memory for comprehensive analysis.
INTERESTING TO KNOW
💰US government secures 15% cut from Nvidia and AMD China sales 💰
Nvidia and AMD have agreed to pay the US government 15% of their revenues from AI chip sales to China in exchange for export licenses. The unusual deal allows Nvidia to sell its H20 accelerators and AMD its MI308 chips to Chinese customers, with the government potentially netting around $1 billion per quarter if sales return to previous levels. President Trump announced the agreement, calling it a "little deal" that balances national security concerns with business interests.
📈 AI startups raid Wall Street for quant talent with multimillion offers 📈
Major AI labs including OpenAI and Anthropic are aggressively recruiting quantitative researchers from top Wall Street trading firms through rooftop mixers and targeted events. Entry-level quants are now receiving multimillion-dollar offers from AI companies, with equity packages replacing traditional Wall Street bonus structures.
Tech recruiter Mike Doonan reports a 40-50% increase in AI startups specifically targeting quant finance talent over the past 18 months, as these professionals bring expertise in algorithm optimization and data processing that directly translates to AI development and safety challenges.

📩 Have questions or feedback? Just reply to this email , we’d love to hear from you!
🔗 Stay connected: