LLMs

Twitter

Facebook

Copy Link

Key Facts

Year2018

Key PlayersOpenAI, Google, Anthropic

Top News

Wikimedia Foundation Takes Middle Ground in OpenAI Copyright Case, Warns of Open Web Risks

The Wikimedia Foundation filed an amicus brief in The New York Times' lawsuit against OpenAI and Microsoft, arguing that AI model training constitutes fair use but chatbots reproducing content without attribution infringe copyright. The organization warns that an overly simplified ruling could either stifle innovation or fragment the open web's knowledge flow.

Policy3 sources

Wikimedia Foundation Takes Middle Ground in OpenAI Copyright Case, Warns of Open Web Risks

Policy3 sources

Orange Cyberdefense Partners with Mistral AI to Deploy Large 4 Model for Cybersecurity Operations

Orange Cyberdefense has become the first European managed security services provider to join Mistral AI's design partner programme for Large 4. The partnership will deploy a cybersecurity-focused AI model across penetration testing and breach investigation services, with operations running exclusively in European data centers to maintain data sovereignty.

Technology4 sources

Orange Cyberdefense Partners with Mistral AI to Deploy Large 4 Model for Cybersecurity Operations

Technology4 sources

Mistral AI Releases Le Chonk: Trillion-Parameter Open-Weight Model Claims Lead in Cybersecurity

French AI company Mistral AI has released the weights for its trillion-parameter open-weight AI model, Mistral Large 4, nicknamed Le Chonk. The model targets cybersecurity and agentic coding tasks, claiming to outperform Chinese competitors while using 47% fewer tokens. However, Mistral acknowledges the model could aid real cyberattacks despite built-in safeguards.

Technology14 sources

Mistral AI Releases Le Chonk: Trillion-Parameter Open-Weight Model Claims Lead in Cybersecurity

Technology14 sources

DeepSeek and Moonshot AI Allegedly Trained Latest Models Using OpenAI's GPT-6 Astra

Chinese AI developers DeepSeek and Moonshot AI reportedly used outputs from OpenAI's GPT-6 Astra to train their latest models through distillation, according to The Information. The practice has intensified concerns over AI distillation and US-China tech competition as leading labs implement safeguards against model-copying.

Technology3 sources

DeepSeek and Moonshot AI Allegedly Trained Latest Models Using OpenAI's GPT-6 Astra

Technology3 sources

RTX Spark Laptops Excel at Local AI Coding But Memory Constraints and Price Tags Raise Questions

Nvidia's RTX Spark platform delivers impressive local AI performance in laptops like the Lenovo Yoga 9n, handling complex coding tasks and running large language models on-device. But real-world testing exposes critical memory limitations, compatibility challenges, and pricing concerns that potential buyers need to understand before investing $3,000-plus in this AI hardware platform.

Technology7 sources

RTX Spark Laptops Excel at Local AI Coding But Memory Constraints and Price Tags Raise Questions

Technology7 sources

Moonshot AI Limits Hacking Ability in Kimi K3.1 Pro Amid Growing AI Cyberattack Concerns

Beijing-based Moonshot AI deliberately reduced the cyber-offensive capabilities of its new Kimi K3.1 Pro model before releasing its weights on November 4. The company removed cyber-related data from training while retaining coding skills, scoring 81% versus 90.5% on internal benchmarks without safeguards.

Technology5 sources

Moonshot AI Limits Hacking Ability in Kimi K3.1 Pro Amid Growing AI Cyberattack Concerns

Technology5 sources

Moonshot AI Launches Kimi K3.1 Pro: 3 Trillion Parameter Model Outperforms GPT-6 Astra

Moonshot AI unveiled Kimi K3.1 Pro, a 3 trillion parameter open-weight AI model featuring a 1-million-token context window. The model outperforms GPT-6 Astra on multiple benchmarks including Terminal-Bench 4.0, scoring 65.7% versus Astra's 58.2%. Full model weights release is scheduled for November 4.

Technology4 sources

Moonshot AI Launches Kimi K3.1 Pro: 3 Trillion Parameter Model Outperforms GPT-6 Astra

Technology4 sources

DeepSeek and MiniMax Clear China AI Export Controls for Global AI Expansion

China approved DeepSeek and MiniMax for overseas deployment under new AI export restrictions covering 23 technologies. The move signals Beijing's strategy to balance global expansion of its AI sector with national security concerns while keeping intellectual property onshore.

Policy7 sources

DeepSeek and MiniMax Clear China AI Export Controls for Global AI Expansion

Policy7 sources

Railway Cyberattack Exposes Customer API Keys as Attackers Target AI Infrastructure

Cloud hosting platform Railway disclosed a security breach where attackers accessed encrypted environment variables and decrypted sensitive customer data. The company linked the incident to a broader campaign targeting AI inference providers and urged immediate credential rotation across all services.

Technology2 sources

Railway Cyberattack Exposes Customer API Keys as Attackers Target AI Infrastructure

Technology2 sources

Model Routing Slashes AI Inference Costs While Maintaining Quality

Organizations running AI models face mounting inference costs, but model routing offers a solution. By dynamically selecting the most appropriate AI model based on task complexity, companies achieve 20-95% cost reductions. Major vendors like Databricks, NVIDIA, and Snowflake now offer routing solutions, though teams must avoid the Pareto trap that can turn savings into customer retention losses.

Technology2 sources

Model Routing Slashes AI Inference Costs While Maintaining Quality

Technology2 sources

Eternal Unveils Dark.ai, an End-to-End Encrypted AI Assistant for Sensitive Conversations

Zomato and Blinkit parent Eternal announced Dark.ai, a privacy-focused AI platform with end-to-end encryption that prevents even the company from reading user conversations. Founder Deepinder Goyal positions the AI assistant as a solution for power users including founders, researchers and enterprises who need to discuss sensitive information without concerns about data retention or access.

Technology7 sources

Eternal Unveils Dark.ai, an End-to-End Encrypted AI Assistant for Sensitive Conversations

Technology7 sources

KRICT Develops Conversational AI That Lets Researchers Guide Self-Driving Labs Using Natural Language

Korean researchers at KRICT developed conversational AI technology that allows self-driving labs to incorporate researchers' expertise through natural language. The human-in-the-loop optimization framework uses large language models to reduce experimental searches by up to 85%, potentially saving 20 days in catalyst research.

Science2 sources

KRICT Develops Conversational AI That Lets Researchers Guide Self-Driving Labs Using Natural Language

Science2 sources

Google AMIE transitions from simulation to real-world clinical practice with 100-patient study

Google's Articulate Medical Intelligence Explorer completed its first real-world clinical trial at Beth Israel Deaconess Medical Center, where 100 patients interacted with the AI agent before primary care appointments. The study published in The Lancet shows AMIE enabled safe clinical conversations and helped physicians prepare for visits, though researchers emphasize it's not ready for standalone diagnosis or patient management.

Health2 sources

Google AMIE transitions from simulation to real-world clinical practice with 100-patient study

Health2 sources

Inception Mercury 3 Diffusion Model Claims 8X Speed Advantage Over GPT-6 Luna and Claude Haiku 5.5

Redwood City AI lab Inception released Mercury 3, a diffusion-based large language model that generates more than 1000 tokens per second on Nvidia chips. The model outperforms GPT-6 Luna and Claude Haiku 5.5 across benchmarks while running four to eight times faster, with a one million token context window and launch pricing of $0.10 per million input tokens.

Technology2 sources

Inception Mercury 3 Diffusion Model Claims 8X Speed Advantage Over GPT-6 Luna and Claude Haiku 5.5

Technology2 sources

OpenAI Claims Breakthrough in COBOL Code Crisis with 98% Migration Accuracy Using GPT-6.1 Sol

OpenAI announced its GPT-6.1 Sol model achieved 97.9% accuracy in migrating COBOL programs to modern languages like Python and Java while preserving behavior. The breakthrough addresses the decades-old challenge of modernizing legacy systems that power 95% of ATM transactions and 40% of online banking systems, though experts caution production deployment still requires extensive testing.

Technology5 sources

OpenAI Claims Breakthrough in COBOL Code Crisis with 98% Migration Accuracy Using GPT-6.1 Sol

Technology5 sources

Prev12 3 4…33 Next