Meta's Muse Spark is the company's first AI model since the social media giant's previous initiatives, and analysts are optimistic about its potential.
Cohere's Transcribe model is designed for tasks like note-taking and speech analysis, supporting 14 languages and optimized for consumer-grade GPUs, making it accessible for self-hosting.
Kroiss led pretraining, which helps train the company's AI models on large datasets, and reported directly to Musk. He also worked on improving xAI's coding models alongside Zhang, who left earlier in March.
PolarQuant is doing most of the compression, but the second step cleans up the rough spots. Google proposes smoothing that out with a technique called Quantized Johnson-Lindenstrauss (QJL).
In 2025, we rebuilt the foundations of our AI program [and] over the coming months, we're going to start shipping our new models and products, and I expect us to steadily push the frontier over the course of the new year. Our world-class recommendation systems are already driving meaningful growth across our apps and ads business, but we think that the current systems are primitive compared to what will be possible soon. Today, our systems help people stay in touch with friends, understand the world, and find interesting and entertaining content.
Anthropic has doubled the amount of VC funding it aims to raise, the FT reports, increasing the target from $10 billion to $20 billion. The round, expected to close soon, will give the company a valuation of $350 billion, sources told the FT. Anthropic, which makes the popular AI Claude and the equally popular Claude Code, decided to double its funding target due to booming investor interest.
TL;DR: 1min.AI gives you lifetime access to many of today's most popular AI models and creative tools text, images, audio, video, and PDFs for a single $74.97 payment (reg. $540) through Jan. 31. Using AI casually might have worked a year ago, but today, professionals need more than one chatbot and a few prompts. The 1min.AI Advanced Business Plan is built for people who want real, repeatable results from AI, without juggling multiple subscriptions.
The stock market is quickly waking up to the fact that is the leading AI software company today. It has overtaken and is less than $200 billion away from overtaking as the world's second-largest company. Overtaking the AI hardware leader, will take a bit more work, as it has a $755 billion lead over Google. Fortuitously for Google, that number is a drop in the bucket these days for Wall Street.
The pace of change in the burgeoning generative AI world is blisteringly fast. It's often hard to keep up with everything, even if it's your full-time job. Readers tell me that one area they find particularly confusing is the wide array of poorly-named AI models. What in the heck is the difference between GPT-5.1, Opus 4.5, Gemini 3, etc.? And why would you use one over the other?
November was the month AI went full gladiator mode: three frontier labs released their best models within a week, Google reclaimed the throne with Gemini 3, and open source proved it can win Olympic gold medals in mathematics. Meanwhile, the agent revolution became official doctrine at Microsoft and Google, and China's AI ecosystem hit escape velocity with 10 million app downloads in seven days.
The move is a big one for Apple, which has traditionally relied on its own technology but plans to use Google's model as a temporary solution until its own AI becomes powerful enough, including to power a slate of upcoming features for the voice assistant. The custom AI model's 1.2 trillion parameters - a measure of the software's complexity and capability - would far exceed the level of Apple's current models.
You could almost call it too much of a good thing. AI models have gone open source -- en masse -- to the tune of millions now competing for a place in our applications. There are echoes of the open sourcing of vital enterprise software and tools, with the opening of operating systems, databases, development platforms, and utilities at no cost. There is a difference this time, however. The AI world has virtually exploded into a mind-blowing array of models that handle everything from neural networks to risk-management tools.
Privacy-focused consumer tech company DuckDuckGo launched a subscription plan last year that bundled a VPN service, personal information removal, and identity theft restoration. The company said Thursday that the subscription now gives users access to the latest AI models through Duck.ai without paying extra. The Duck.ai chatbot is free to use, and users get access to models like Anthropic's Claude 3.5 Haiku, Meta's Llama 4 Scout, Mistral AI's Mistral Small 3 24B, and OpenAI's GPT-4o mini.
The premium plan called ChatGPT Plus guarantees access during blackout periods and includes access to OpenAI's most advanced models, making it attractive for superusers.
OpenAI's GPT-5 model marks a significant upgrade in coding and creative writing. However, it still lacks crucial elements needed for true AGI, including continuous learning.
