the-decoder7d
Mistral AI's Le Chat update brings up to 1,000 words per second responses and new mobile apps
Mistral AI has rolled out a major update to its AI assistant Le Chat, adding new features and mobile apps while revamping its pricing structure. One of the most notable additions is "Flash Answers," ...
the-decoder4d
Want to understand ChatGPT? Watch Andrej Karpathy's explanation of how LLMs work
If you want to understand AI, there's one video you need to watch this week. Andrej Karpathy, formerly of OpenAI and Tesla, has released what might be the clearest explanation yet of how Large ...
the-decoder6d
OpenAI co-founder John Schulman's brief stint at Anthropic comes to an end
John Schulman, who helped create ChatGPT's underlying technology at OpenAI, has left Anthropic after just five months with the company. As one of OpenAI's co-founders, he played a key role in ...
the-decoder6d
AlphaGeometry2: Deepmind AI outperforms math Olympians at geometry tasks
The latest version of Deepmind's AlphaGeometry system can solve geometry problems better than most human experts, matching the performance of top math competition winners. AlphaGeometry2 solves 84% of ...
the-decoder9d
ByteDance's AI can animate both real people and cartoon characters from a single image
Researchers at TikTok parent company ByteDance have presented OmniHuman-1, a new framework for generating videos from image and audio samples. The new system from TikTok's parent company turns still ...
the-decoder4d
Pika Labs launches new "Pikadditions" video inpainting feature
Pika Labs' new "Pikadditions" tool allows users to seamlessly insert objects into existing videos. The process is simple: upload a video and an image, and the system places the object from the image ...
the-decoder4d
OpenAI introduces EU data processing for ChatGPT Enterprise, Edu, and API customers
OpenAI is expanding its services in Europe by letting companies store their data in EU-based data centers. The change applies to ChatGPT Enterprise, ChatGPT Edu, and the company's API platform. For ...
the-decoder7d
Hugging Face works to replicate OpenAI's Deep Research capabilities with open-source AI agent
A team at Hugging Face, led by chief researcher Thomas Wolf, has created an open-source version of OpenAI's Deep Research system in 24 hours. According to the Hugging Face blog, they aim to make the ...
the-decoder7d
Robotics startup Figure AI drops OpenAI because LLMs are 'getting smarter yet more commoditized'
On X, Figure CEO Adcock explains that "LLMs are getting smarter yet more commoditized" and have "quickly become the smallest piece of the puzzle" for Figure AI. The much harder part, he says, is ...
the-decoder7d
Scientists see potential in AI, but hesitate to embrace it, survey finds
A major international survey shows that while researchers recognize AI's potential in science, most haven't integrated it into their daily work yet. According to a new survey of nearly 5,000 ...
the-decoder4d
Meta's new MILS system teaches LLMs to handle multimedia data without specialized training
Meta AI researchers and their academic partners have developed a system that teaches large language models to handle images, video, and audio without specialized training. Called MILS (Multimodal ...
the-decoder10d
OpenAI's Deep Research aims to compress hours of online research into minutes
OpenAI has launched "Deep Research," a new Agent feature for ChatGPT that aims to streamline complex research tasks. Initially available to Pro users, the tool promises to compress hours of online ...