DeepSeek takes aim at Anthropic with new model
tynmagazine.comDeepSeek released an experimental multimodal model it claims brings its performance close to Anthropic's Opus 4.8 on various benchmarks.
Protoface Launches Protoface Studio and API for AI Video Creation
freep.comProtoface Studio gives creators unlimited access to frontier video models including Google's Veo and MiniMax H3 API, providing one-cent-per-second pricing for AI video creation tools.
Alibaba's Wan3.0 Generates 30-Second AI Videos From Documents
eweek.comArticle about Alibaba's new AI video model Wan3.0 that can generate 30-second videos from documents, targeting enterprise teams with lower prices and expanded capabilities in APAC region.
Alibaba launches AI video model Wan 3.0 after Qwen beats Gemini and ChatGPT in coding
msn.comAlibaba has launched Wan3.0, a new AI video model that can turn text, images, audio, videos and even PDFs into 30-second videos from documents, targeting enterprise teams with lower prices.
OpenAI says upcoming model is so capable it requires stronger guardrails
msn.comOpenAI has determined that one of its upcoming models requires extra safety layers during development and release due to its advanced capabilities.
DeepSeek Vision Model Challenges Gemini on Cost, Trails on Speed
techgig.comDeepSeek's V4 Flash Vision Exp model is challenging Google's Gemini 3.7 Flash with impressive accuracy and significantly lower costs, though it trails in speed performance according to this comparison analysis.
I gave Claude Code a deny list, and stopped watching every command it tried to run
xda-developers.comThis article discusses how the author implemented a deny list for Claude Code, an AI coding assistant that checks permission rules in order: deny first, then ask, and finally allow. The tool now respects these restrictions without constant supervision.
如何看待 MiniMax 闫俊杰提出「最小化推理成本，最大化智能」? 这会改变大模型的「斩杀线」吗?
msn.cnMiniMax's Yan Junjie proposes a strategy to minimize inference costs while maximizing intelligence, potentially changing the cost-performance threshold for large language models. The article discusses M3.1 model goals targeting one-third of initial unit reasoning costs.
MiniMax H3 的 24h AI 直播火了，但我们搓出了更炸裂的东西！
news.qq.comAn FAL engineer integrated MiniMax H3 Max into a Twitch livestream, creating viral 24-hour AI content. The article discusses the model's capabilities and potential for extended generation tasks.
Perplexity CEO announces rollout of hybrid compute feature for Mac application
msn.comPerplexity introduced hybrid compute for all users of its Mac application, enabling tasks to be processed with a combination of cloud-based and local AI models.
Qualcomm rival MediaTek jumps 10% after $3.5 billion Nvidia AI chip deal
msn.comMediaTek shares jumped 10% after announcing a $3.5 billion deal with Nvidia for AI chips, as the two companies will work on integrating Nvidia technology with MediaTek's custom AI chip business and other areas.
OpenAI: Apple's trade secret theft accusation is baseless
msn.comOpenAI argues in a court filing that Apple's accusations of trade theft against the AI company are baseless, defending its position on proprietary model development.
Apple's accusations are making things uncomfortable for OpenAI
msn.comApple accuses OpenAI of destroying evidence in its trade secrets case, claiming a former engineer sent destruction orders related to AI model development.
Judge rules Trump administration illegally punished AI firm Anthropic
msn.comA judge ruled that the Trump administration illegally punished Anthropic for protected speech, barring enforcement of rules aimed at cutting off the AI company from federal government contracts.
OpenAI to launch new model with 'stronger safeguards' after hack
msn.comOpenAI is preparing to release its new Astra model with stronger safeguards following a security hack, marking an important development in AI safety and deployment.
Anthropic's $1.5B piracy settlement is back to haunt it
politico.comAnthropic faces renewed legal challenges over allegedly using copyrighted works to train its models without permission, with opponents taking a more targeted litigation approach.
AGI: Can India turn disruption into opportunity?
msn.comWith AGI potentially closer than experts suggest, India must prepare for both the opportunities and disruptions that will come from this technological advancement. The article discusses economic implications of AGI development in India.
Federal Judge Rules Pentagon's AI Blacklist Violated The Constitution
yahoo.comA federal judge ruled that the Pentagon violated the Constitution by blacklisting Anthropic as a supply chain risk in retaliation for their AI safety work.
Creepy 'Cat in the Hat' AI trend targets Tennessee schools. What parents should...
yahoo.comThe Tennessee Department of Safety is tracking a "Cat in the Hat" AI trend that targets schools, prompting concerns for parents.
Viral Cat in the Hat AI trend sparks safety warnings for schools and students
yahoo.comA viral "Cat in the Hat" trend using unsettling AI-generated videos and images has prompted safety warnings for schools and students.