News

the-decoder3d
AWS reportedly faces customer frustration over Anthropic usage limits
Customers are expressing frustration with Amazon Web Services over constraints in its AI platform Bedrock, according to a report from The Information. Despite AWS investing in Anthropic, the company ...
the-decoder3d
So-called reasoning models are more efficient but not more capable than regular LLMs, study finds
RLVR helps repetition, not generalization AI researcher Nathan Lambert describes the findings as consistent with expectations. "This isn’t a new intuition," he writes, "but a nice new set of results." ...
the-decoder3d
Alibaba’s VACE AI model aims to become the universal tool for video generation and editing
Scientists at Alibaba Group have introduced VACE, a general-purpose AI model designed to handle a broad range of video generation and editing tasks within a single system. The model’s backbone is an ...
the-decoder3d
ChatGPT Search reaches approximately 41 million monthly users in the EU
OpenAI’s ChatGPT Search recorded approximately 41.3 million monthly users in the European Union over the six-month period ending in March 2025, according to the company’s own data. The figure ...
the-decoder4d
Google’s AI Overviews are quietly draining clicks from top sites, new data shows
New analysis from Ahrefs shows Google’s "AI Overviews" are driving down clicks to top-ranked websites by over 34%, directly contradicting Google’s own claims. A recent analysis from Ahrefs suggests ...
the-decoder5d
Grok 3 Mini turns up the heat as AI price wars push model costs even lower
xAI is making a push on efficient AI with the release of Grok 3 Mini, its newest language model. Both Grok 3 and its Mini sibling are available through the xAI API. The Grok 3 family currently ...
the-decoder5d
The next leap in AI depends on agents that learn by doing, not just by reading what humans wrote
Richard S. Sutton's "Bitter Lesson" lays out a hard truth at the heart of modern AI: Not the clever injection of human knowledge, but scalable learning and search algorithms are what deliver lasting ...
the-decoder5d
OpenAI's o3 achieves near-perfect performance on long context benchmark
With support for up to 200,000 tokens, o3 is the first model to achieve a perfect 100 percent on the Fiction.live benchmark using 128,000 tokens—that’s roughly 96,000 words. For any language model ...
the-decoder5d
Students delegate higher-level thinking to AI, Anthropic study finds
A new study from Anthropic examines how university students are using its language model Claude in daily academic work. The analysis reveals discipline-specific usage patterns and raises concerns ...
the-decoder6d
BitNet: Microsoft shows how to put AI models on a diet
BitNet b1.58 2B4T is a new language model from Microsoft designed to operate with minimal energy and memory usage. Unlike conventional language models that rely on 16- or 32-bit floating point numbers ...
the-decoder6d
GPT-4o makes beautiful images but fails basic reasoning tests, UCLA study finds
Despite recent progress in image generation quality, the empirical analysis reveals notable weaknesses in how GPT-4o handles complex prompts. Researchers evaluated the model across three categories: ...
the-decoder6d
Researchers introduce COLORBENCH to test color understanding in vision-language models
According to the researchers, the results reveal fundamental weaknesses in color perception—even among the largest models currently available. Color plays a central role in human visual cognition and ...