the-decoder6d
Perplexity announces new assistant for Android smartphones
Perplexity is stepping into Google's territory with a new AI assistant for Android that can control apps and handle tasks on its own. The move puts the startup in direct competition with Google's ...
the-decoder4d
OpenAI's ChatGPT gets a major Canvas upgrade with HTML and React code rendering
The updates include support for the new o1 model, which users can now select from the model picker, though this option is limited to Pro, Plus and Team subscribers. One of the most practical additions ...
the-decoder4d
Deepseek puts pressure on Meta with open source AI models at a fraction of the cost
In recent weeks, Chinese AI startup Deepseek has shown that cutting-edge AI development doesn't require massive budgets, putting pressure on established AI labs. Meta CEO Mark Zuckerberg is doubling ...
the-decoder4d
Chinese OpenAI o1 challenger Kimi k1.5 now available as free web version
Moonshot AI's latest reasoning model, Kimi k1.5, is now available to everyone through Kimi.ai. The model now works in English too, though the company says it's still fine-tuning the language support.
the-decoder3d
'Inscrutable Wizards': How Chinese AI startup Deepseek is making Silicon Valley look slow
Deepseek out-acclerates Sillcon Valley accelerators: The company's latest model, Deepseek-V3, performs better than leading commercial AI systems in benchmark tests, according to independent ...
the-decoder3d
New Tencent AI model Hunyuan3D 2.0 turns 2D images into detailed 3D objects
Tencent has released version 2.0 of Hunyuan3D, an open-source generative AI system that creates textured 3D models from regular images. The system splits the work between two specialized components: ...
the-decoder6d
OpenAI's Operator and Computer-Using Agent bring autonomous AI agents closer to reality
OpenAI has just launched Operator, an AI assistant that can navigate the web on its own. The tool, currently only available to US ChatGPT Pro subscribers, represents a step toward AI assistants that ...
the-decoder5d
Frontier models fail hard at "Humanity's Last Exam" but experts question if it matters
An international research team has developed a new benchmark that reveals the current limitations of LLMs. Even the most advanced models fail at 90 percent of the tasks - for now. The test, called ...
the-decoder4d
Anthropic adds source citations to Claude models
Anthropic has rolled out "Citations," a new API feature that helps Claude automatically add source references to its responses, making AI-generated content more reliable and trustworthy. The company ...
the-decoder3d
Former Beatle Paul McCartney warns of 'rip-off' AI copyright changes
The UK government is looking to change its laws to let AI companies train their models using online content from creators, unless those creators specifically opt out of the system. In a BBC interview, ...
the-decoder4d
Meta's Llama stack makes it easier to develop typical LLM use cases
Meta has launched a stable API release of its development platform called Llama-Stack 0.1.0 to help companies build AI applications using Llama models. The platform provides standardized building ...
the-decoder6d
"Nerd sniping" and "think less" attacks emerge as AI models get more time to reason
A new study by OpenAI shows that AI models become more robust against manipulation attempts if they are given more time to "think". The researchers also discovered new methods of attack. A recent ...