the-decoder

New Deepseek technique balances signal flow and learning capacity in large AI models

DeepSeek researchers have developed a technique that makes training large language models more stable. The approach uses mathematical constraints to solve a well-known problem with expanded network ...
the-decoder

AI benchmarks are broken and the industry keeps using them anyway, study finds

AI benchmarks are pretty unreliable, Epoch AI finds: Scores shift a lot based on prompt wording, temperature settings, or which testing tool is used. The API provider causes the widest swings in ...
the-decoder

Claude creator Anthropic reportedly hits $350 billion valuation as it raises another $10 billion

Anthropic, the company behind the Claude chatbot, is raising $10 billion at a valuation of $350 billion - nearly double its $183 billion valuation from just four months ago. Singapore's sovereign ...
the-decoder

Dr. ChatGPT is now officially in, as OpenAI launches dedicated health section for its chatbot

OpenAI has launched ChatGPT Health, a dedicated section within ChatGPT where users can connect medical records and wellness apps to get help understanding lab results and preparing for doctor visits.
the-decoder

AMD unveils AI accelerators and laptop chips at CES 2026

AMD presented the complete Instinct MI400 portfolio for data centers at CES 2026. OpenAI will receive the first MI400 chips this year based on a deal that is expected to bring AMD billions in ...
the-decoder

Matthias Bastian

Read full article about: Microsoft and Stripe bring shopping checkout directly into Copilot chat for US users In a glorious AI future, you'll order pizza directly from Excel. Microsoft and Stripe are ...
the-decoder

Maximilian Schreiner

Read full article about: More than five percent of ChatGPT messages worldwide are about health More than five percent of all messages sent through ChatGPT worldwide deal with health topics. According ...
the-decoder

Meta Neuroscientist King: "Some of the concepts like reasoning may need to be re-evaluated"

The neuroscientist Jean-Rémi King leads the Brain & AI team in Meta’s AI division. In an interview with The Decoder, he discusses the connection between AI and neuroscience, the challenges of ...
the-decoder

Deepfakes: How it all began - and where it could lead us

AI-faked images and videos - so-called deepfakes - have developed rapidly in recent years. Here, we trace the history and describe the most important milestones. What do all the people in the cover ...
the-decoder

OpenAI co-founder Ilya Sutskever believes AI will shape everyone's life "whether you like it or not"

Former OpenAI chief scientist Ilya Sutskever spoke at the University of Toronto, urging people to prepare for the major shifts artificial intelligence will cause. He said that AI could soon take on ...
the-decoder

Adobe Photoshop integrates Google's image AI "Nano Banana"

Adobe is bringing Google's new image AI, known as "Nano Banana" (officially Gemini 2.5 Flash Image), to Photoshop as an optional tool. The model is designed for editing existing images with a high ...
the-decoder

ChatGPT quietly switches to a stricter language model when users submit emotional prompts

OpenAI is piloting a new safety feature in ChatGPT that automatically routes sensitive or emotional user requests to a stricter language model with more limitations. Technical analysis by one user ...