Scaling AI for everyone
OpenAI secures $110B in new funding at $730B pre-money valuation: $30B from SoftBank, $30B from NVIDIA, $50B from Amazon. Major capital round to scale AI deployment globally.
Every article scored by Claude on two independent axes: signal (useful info) and hype (clickbait). Filtered before you read.
OpenAI secures $110B in new funding at $730B pre-money valuation: $30B from SoftBank, $30B from NVIDIA, $50B from Amazon. Major capital round to scale AI deployment globally.
Vercel releases coordinated security patch for Next.js addressing 13 vulnerabilities: auth bypass via App Router, dynamic route parameter injection, cache poisoning, DoS in React Server Components (CVE-2026-23870), and XSS. Immediate upgrade mandatory for all affected users.
Anthropic raises $965B Series H and launches Opus 4.8 with Dynamic Workflows and ultracode. Major funding expansion and new model capabilities.
Release of llm-anthropic 0.25.1: adds Claude Opus 4.8 model, -o fast 1 option for fast mode (enabled organizations), and default max_tokens now matches each model's maximum output instead of 8192.
Anthropic raises $65 billion in Series H at a $965 billion valuation. Annualized revenue reaches $47 billion according to CFO Krishna Rao. The company will invest in safety research, computing capacity, and expanding its Claude product lineup.
ITBench-AA, a new benchmark from Artificial Analysis and IBM, evaluates frontier models on agentic enterprise IT tasks. Top models (Claude, GPT-4, Gemini) score below 50%, exposing significant gaps in automating complex IT workflows.
Meta releases code and checkpoints for SAM 3 (Segment Anything Model 3). Repository includes inference, fine-tuning, and example notebooks for image segmentation.
Impossibility theorem: no feature ranking can be simultaneously faithful, stable, and complete under collinearity. Authors quantify the result for 4 model classes, propose DASH (Diversified Aggregation of SHAP) as resolution, and formally verify 305 Lean 4 theorems. Consequence: 68% of public datasets exhibit attribution instability.
OpenAI's reasoning model disproved a 1946 Erdős conjecture in unit-distance geometry using unexpected algebraic number theory tools. Fields Medalist Tim Gowers calls it "a milestone in AI mathematics."
OpenAI Whisper is a speech recognition model trained on 680,000 hours of multilingual weakly supervised data. The GitHub repository includes code, pre-trained models, and performance benchmarks across multiple languages and acoustic conditions.
SpaceX signed a Cloud Services Agreement with Anthropic to provide compute capacity on COLOSSUS and COLOSSUS II clusters. Anthropic will pay $1.25 billion per month through May 2029, with reduced fees during May-June 2026 ramp-up. SpaceX uses these resources to train Grok 5.
An OpenAI model disproved a major conjecture in discrete geometry by solving the 80-year-old unit distance problem. This breakthrough marks a milestone in AI-driven mathematics.
Systematic analysis of 40 agent safety benchmarks (2023-2026). Benchmarks exhibit incompatible threat models, fragmented metrics, and inconsistent risk coverage. Concordance test (Kendall's W = 0.10, p = 0.94) reveals no ranking alignment across evaluation dimensions. Releases structured metadata and proposes minimum reporting standards.
Google DeepMind introduces Gemini Omni, a multimodal model processing text, audio, video, and images as native inputs and outputs. The model delivers ultra-low latency and improved performance on reasoning and vision benchmarks.
OpenAI raises $122 billion to accelerate frontier AI development, expand compute capacity, and meet growing demand for ChatGPT, Codex, and enterprise AI solutions.
OpenAI releases GPT-5.4, its most capable and efficient frontier model for professional work, with state-of-the-art coding, computer use, tool search, and 1M-token context window.
OpenAI and AWS announce multi-year strategic partnership worth $38 billion. AWS will provide infrastructure and compute capacity to power OpenAI's next-generation models.
Gemini 2.5 Deep Think achieves gold-medal level performance at the International Collegiate Programming Contest World Finals, demonstrating a major breakthrough in abstract problem-solving capabilities.
OpenAI releases o3 and o4-mini, its most capable models to date with full tool access. o3 marks a leap in reasoning and complex problem-solving capabilities. o4-mini provides a lighter, more accessible alternative.
OpenAI secures $40B funding at $300B post-money valuation to advance AI research, scale compute infrastructure, and support 500M weekly ChatGPT users.
OpenAI releases o3-mini, a compact reasoning model optimized for efficiency. Designed for complex tasks with reduced latency and lower costs, it delivers o3-comparable performance on code and math benchmarks.
Hugging Face reproduces DeepSeek-R1, an open-source reasoning model. Open-R1 provides a fully open alternative to proprietary models, with code, data, and weights publicly available for research and deployment.
Sora, OpenAI's video generation model, is now available at sora.com. It produces videos up to 1080p, maximum 20 seconds, in landscape, portrait, or square formats. Users can generate content from text or remix existing assets.
OpenAI launches Realtime API enabling developers to build fast bidirectional speech experiences. The API supports speech input/output with low latency and native function calling integration.
OpenAI introduces o1, a reasoning model capable of solving complex problems in mathematics, coding, and science. The model uses internal reflection before responding, improving performance on difficult benchmarks.
OpenAI releases o1-mini, a smaller and more cost-efficient reasoning model compared to o1. Designed for complex reasoning tasks with improved cost-performance ratio.
OpenAI makes fine-tuning available for GPT-4o. Users can now customize the model for specific use cases through the API.
OpenAI introduces Structured Outputs in the API. Models now reliably produce JSON outputs that conform to developer-supplied schemas, eliminating parsing errors and improving application reliability.
Meta releases Llama 3.1 in three sizes (405B, 70B, 8B) with multilingual support and extended context. Models support 128k tokens and cover 8 languages. Available open-source via Hugging Face.
OpenAI releases GPT-4o mini, a smaller and cheaper model than GPT-4o. It delivers comparable performance on many tasks while reducing inference costs. The model supports text, vision, and audio.
OpenAI announces GPT-4o, its new flagship model capable of reasoning across audio, vision, and text in real time.
OpenAI makes GPT-4o available to free ChatGPT users alongside new capabilities. The flagship model becomes accessible without paid subscription.
OpenAI releases GPT-4o and expands free ChatGPT access with additional capabilities. The model improves multimodal performance and processing speed.
OpenAI releases ChatGPT and Whisper APIs, enabling developers to integrate conversational AI and speech recognition into applications. The APIs provide programmatic access to ChatGPT's conversation capabilities and Whisper's audio transcription features.
Google releases CodeGemma, a family of code-specialized language models based on Gemma. Available in 7B and 2B sizes with open weights, CodeGemma includes pre-trained and instruction-tuned variants optimized for coding tasks.
OpenAI introduces Sora, a text-conditional diffusion model trained jointly on videos and images of variable durations, resolutions and aspect ratios. Built on a transformer architecture operating on spacetime patches, Sora generates up to one minute of high-fidelity video. OpenAI suggests that scaling video generation models is a promising path toward general-purpose physical world simulators.
OpenAI announces GPT-4 Turbo with 128K context window and lower pricing, Assistants API, GPT-4 Turbo with Vision, and DALL·E 3 API. Multiple developer products released.
OpenAI launches GPTs, custom versions of ChatGPT combining instructions, extra knowledge, and various skills without requiring coding.
Hugging Face announces the release of Falcon 180B, an open-source large language model with 180 billion parameters. The model is available in base and instruction-tuned versions, designed for complex text generation and reasoning tasks.
Meta releases Llama 2, an open-source language model available on Hugging Face. The model comes in multiple sizes and can be used freely for research and commercial applications.