the-decoder3d
Alibaba's Qwen2.5-VL-32B matches larger models with just 32B parameters
Alibaba has unveiled its latest multimodal AI model, Qwen2.5-VL-32B, released under the Apache 2.0 license. Early benchmarks show the model outperforming larger competitors like Gemma 3-27B and ...
the-decoder7d
Grok 3 adds deeper search and AI image editing capabilities
Grok users are getting access to two new features as xAI rolls out an enhanced research tool and image editor. The updates are being introduced gradually across the platform. The new "DeeperSearch" ...
the-decoder10d
Study: Open source and nearly 3,000 developers are the digital foundation of the global economy
A new Harvard Business School study reveals open-source software represents enormous economic value. Without it, companies would spend 3.5 times more on software, with just 5% of developers generating ...
the-decoder16d
Google opens up its AI Deep Research feature to all users for free "a few times a month"
The latest version of Flash Thinking 2.0 brings new capabilities like file uploads and faster processing speeds. For users of Gemini Advanced, the system now handles much larger amounts of information ...
the-decoder10d
EXAONE Deep: LG AI Research releases reasoning models
LG AI Research has introduced new reasoning models (LRMs) named EXAONE Deep. Available in three sizes (2.4B, 7.8B and 32B parameters), the models have been optimized for math, science and programming, ...
the-decoder5d
Tencent's Hunyuan-T1 reasoning model matches OpenAI's o1 capabilities in benchmarks
Following the approach used for all large reasoning models, Tencent relied heavily on reinforcement learning during development, with 96.7 percent of post-training computing power focused on improving ...
the-decoder11d
OpenAI's post-training chief William Fedus departs to launch science startup
William Fedus, who served as Vice President of Post-Training at OpenAI, has announced his departure from the company to pursue AI applications in scientific research. In an internal memo to colleagues ...
the-decoder24d
Google's new AI mode for search might turn the Web into a World Wide Wasteland
Google's experimental AI mode pushes search toward a closed ecosystem, mirroring OpenAI's ChatGPT search and Perplexity while potentially delivering a fatal blow to the open web. The system, built on ...
the-decoder6d
Kai-Fu Lee says OpenAI's Sam Altman "probably not sleeping well" as 01.AI pivots to Deepseek
Former Google China head Kai-Fu Lee is pivoting his AI startup 01.AI to fully embrace Deepseek's open-source models, describing them as an existential challenge to OpenAI's business model. Lee's ...
the-decoder6d
Microsoft Research has developed a new way to feed knowledge into LLMs
Microsoft Research has developed a more efficient way to incorporate external knowledge into language models. The new system, called Knowledge Base-Augmented Language Models (KBLaM), takes a ...
the-decoder11d
AI coding startup Cognition AI behind Devin hits $4 billion valuation
Cognition AI has secured hundreds of millions in new funding at a nearly $4 billion valuation, with venture capital firm 8VC leading the round. The investment doubles the startup's previous valuation, ...
the-decoder29d
GPT 4.5 is "proof that we can continue the scaling paradigm," says OpenAI's Chief Research Officer
OpenAI has presented its largest language model to date. According to Mark Chen, Chief Research Officer at OpenAI, GPT 4.5 shows that the scaling of AI models has not yet reached its limits. OpenAI ...