openai OpenAI's o1 Model Exposes Flaws in Compute-Based AI Regulation OpenAI's new "reasoning" model o1 shows that AI can improve without massive computing power, challenging laws like California's SB 1047 that tie safety rules to compute thresholds.
fal.ai Fal.ai Raises $23M to Scale Generative Media Infrastructure for Developers Developer-focused AI platform Fal.ai secures $23 million in combined seed and Series A funding to expand its high-performance infrastructure for generating images, video, and audio. The startup currently serves half a million developers and achieves an annual run rate nearing $10 million.
openai OpenAI Debuts o1 Reasoning Model With Mixed Early Results OpenAI releases its highly anticipated o1 model that "thinks" before answering, offering improved reasoning for complex problems at a much higher cost. However, it currently lacks the speed and multimodal features of GPT-4o.
openai OpenAI Launches o1 Models Capable of Self-Fact-Checking Through Extended Reasoning OpenAI introduces a new family of AI models called o1 that spend more time "thinking" to fact-check themselves and solve complex problems. The release currently faces high costs, strict rate limits, and missing features like web browsing.
google Google Co-Founder Sergey Brin Returns to Work Daily on AI Development Sergey Brin reveals he works at Google "pretty much every day" to focus on artificial intelligence, calling the current progress too exciting to miss. He also expresses doubt that training advanced AI requires endlessly scaling up computing power.
apple Apple Brings AI Translation and Smart Widgets to watchOS 11 Apple reveals new AI-powered features for watchOS 11, including a standalone Translate app and smarter widgets. The updates leverage on-device machine learning to enhance the Apple Watch experience.
nvidia NVIDIA StormCast AI Model Delivers More Accurate Extreme Weather Forecasts NVIDIA introduces StormCast, a generative AI model that improves extreme weather predictions by providing high-resolution, hourly updates. The new system outperforms traditional forecasting methods to support better disaster planning.
google Google DeepMind AI Achieves Silver Medal Level in International Math Competition Google DeepMind introduces AlphaProof and AlphaGeometry 2, two AI systems that successfully solve complex mathematical problems requiring advanced reasoning. The breakthrough allows the AI to earn the equivalent of a silver medal at the prestigious International Mathematical Olympiad.
mistral Mistral Unveils Large 2 to Rival Meta and OpenAI Models Mistral introduces its new Large 2 AI model, claiming it matches or beats competitors in coding and math while using fewer parameters. The model also focuses on reducing hallucinations and offering broad multilingual support.
meta Meta Unveils Llama 3.1 to Rival Top Closed AI Models Meta releases its largest open-source AI model to date, Llama 3.1 405B, which directly competes with leading closed models like GPT-4o and Claude 3.5 Sonnet. The updated model family also features an expanded context window and new multilingual capabilities.
google Google Releases Gemma 2 Open Source Models That Rival Larger AI Google DeepMind open-sources Gemma 2, a family of small language models that use knowledge distillation to outperform competitors of similar size and rival models twice as large.
meta Meta Unveils Llama 3.1 405B as Largest Open-Source AI Model Meta launches Llama 3.1 405B, a massive 405-billion-parameter open-source language model backed by major cloud providers. The model features a 128k token context window, multilingual support, and advanced reasoning tools.
openai OpenAI Introduces CriticGPT to Catch Errors in AI Outputs OpenAI unveils CriticGPT, a new model designed to review and correct mistakes made by other AI systems like ChatGPT. This approach leverages AI to improve AI, aiming to reduce hallucinations and boost overall accuracy.
openai OpenAI Builds CriticGPT to Catch Code Errors Human Trainers Miss OpenAI introduces CriticGPT, a new AI model designed to help human trainers spot code mistakes in ChatGPT responses. This tool addresses the growing challenge of human workers struggling to evaluate increasingly advanced AI outputs.
openai OpenAI Launches CriticGPT to Catch Hidden Errors in ChatGPT Code OpenAI unveils CriticGPT, a GPT-4-based model designed to help human trainers spot mistakes in ChatGPT's code outputs. The new tool significantly improves the accuracy of AI evaluations by overcoming the limitations of manual reviews.