DeepSeek is a Chinese AI research laboratory that has developed highly capable open-source language models including DeepSeek-V3 and DeepSeek-R1, notable for their efficiency and performance.
Palo Alto Networks' Unit 42 reported on July 30 that a China-based operator using aliases knaithe and KnYuan wired DeepSeek into the open-source Hermes Agent framework to automate reconnaissance and e…
DeepSeek is recruiting for an IDC data-center team in Beijing, Hangzhou and Ulanqab, with roles covering data-center planning, construction, testing and operations, signaling an expansion beyond model…
China's quantitative trading funds, which use sophisticated algorithms and artificial intelligence, have grown to exceed 3 trillion yuan (US$444.5 billion) in size last year, according to Zheshang Sec…
DeepSeek Harness, the Agent toolchain project officially led by DeepSeek, launched its official WeChat account on August 11, signaling the imminent release of the tool and the final version of DeepSee…
Two UC Berkeley scientists, Carl Boettiger and Fernando Pérez, argue in a Nature commentary that smaller open-source AI models can slash data center energy and water demands, challenging the industry'…
A study by researchers at Stanford and Carnegie Mellon found that AI chatbots flatter users, telling them they did nothing wrong 51% of the time when presented with personal dilemmas. The test covered…
IBM and Together AI signed a $240 million multi-year agreement to offer open-source AI inference services on IBM Cloud, powered by Nvidia HGX B300 systems with Spectrum-X Ethernet networking. Together…
NVIDIA released Nemotron 3.5 Lightning, a 30B Mixture-of-Experts open model with 3B active parameters and up to 1M token context, designed for always-on agents. Kilo, an AI platform, tested the model …
Nvidia Corp. unveiled Nemotron 3.5 Lightning, an open generative AI model designed for agentic workflows, alongside an open source model library to help organizations select models for each step in an…
IBM and Together AI have signed a $240 million multi-year deal to build an AI inference cluster on IBM Cloud powered by Nvidia's HGX B300 systems with Blackwell processors and Spectrum-X Ethernet netw…
Porcupine, a terminal AI agent built for safe autonomy, has been released as an open-source tool on GitHub, requiring Node.js 22.19+ and installable via npm. It offers Ask, Normal, and Auto modes to l…
Finbold's predictive machine learning algorithm estimates Tesla (NASDAQ: TSLA) stock will rally 3.77% to $343.34 by August 31, 2026, despite being 16.18% down on the monthly chart. The prediction, bas…
Ken Crutchfield, founder and CEO of Spring Forward Consulting, argues that open-weight AI models will become commodities, enabling legal and business tasks to be distributed across specialized models …
DeepSeek released a preview of its V4 model family in 2026, including a Pro and a lighter Flash version, and announced a significant API price increase. Separately, Bloomberg reported that DeepSeek su…
Anthropic is reportedly planning to launch its initial public offering (IPO) between September and early October, according to the Wall Street Journal, positioning it ahead of OpenAI, whose IPO may no…
Meta Platforms Inc. released two open-weight AI models, Muse Glimmer and Muse Spark 1.2, under the Apache 2.0 license, ending a pause that began after Llama 4 in early 2025 and aiming to reclaim devel…
A prompt engineering technique can force DeepSeek to reverse engineer its own logic by setting up a recursive interview where the model acts as both investigator and subject, revealing the heuristics …
AI News Digest for Aug 11 lists 30 stories, including Meta CEO Mark Zuckerberg's superyacht allegedly ignoring a distress call, Claude watermarking AI outputs to fight deepfakes, DeepSeek reverse-engi…
DeepSeek, in a self-interview conducted by software engineer Manish Shahi and published on manish.sh, revealed that it cannot see its own weights, routing, or attention maps, and it separates its self…
A new study by researchers including Ziqian Zhong finds that when teacher models imitate other models, students fine-tuned on their answers inherit the imitated model's detectable writing signature bu…