News reporting focused on AI and machine learning, covering the companies behind these technologies, their real-world applications, and the ethical concerns they raise. This includes areas like generative AI (large language models, text-to-image and video), speech tech, and predictive analytics.
Patreon is taking a tougher stance against AI companies that scrape creator content for training. The membership platform is now working with Cloudflare to actively block AI training bots, rather than just asking them to stop via robots.txt files. The move comes as AI scraping has become more sophisticated and widespread.
Norway's National Library is developing a sovereign large language model for the Norwegian language, using 2 petabytes of Huawei OceanStor Dorado all-flash storage in its AI training pipeline. The project addresses the lack of commercial Norwegian LLMs and highlights challenges in moving data from archival storage to high-performance AI systems.
Amp, a startup in Menlo Park, California, has secured more than $1.3 billion to build a global pool of specialized computer chips for AI training. Founded by Anjney Midha, the company aims to provide access to computing power for startups, universities, and others excluded by tech giants like Amazon, Google, Anthropic, and OpenAI. Investors include Andreessen Horowitz and Y Combinator, with partners such as Periodic Labs and ElevenLabs joining the effort.
NVIDIA has opened Multipath Reliable Connection (MRC), a new RDMA transport protocol optimized for Spectrum-X Ethernet hardware, to the industry via the Open Compute Project. Deployed by OpenAI, Microsoft, and Oracle in major AI factories, MRC improves throughput, load balancing, and availability for large-scale AI training. It enables high GPU utilization by distributing traffic across multiple paths and handles failures in microseconds.
Meta has started using data from its employees' computer activity to train AI agents. The company installed a tool called Model Capability Initiative on US-based workers' machines. It captures mouse movements, clicks, keystrokes, and screenshots to help AI models mimic human computer interactions.
Meta is deploying surveillance software on US employees' computers to record mouse movements, clicks, and keystrokes. The system, known as Model Capability Initiative, aims to train AI agents for independent task handling. Officials state the data stays separate from performance evaluations amid plans for workforce reductions.
Meta plans to collect data on employee mouse movements and keystrokes to improve its AI models. The company will use an internal tool to capture these interactions on specific applications. A spokesperson confirmed safeguards protect sensitive content, with data used only for training purposes.
Clarifai has removed 3 million photos supplied by OkCupid that were used to train its facial recognition technology, according to reports. The action follows an FTC settlement with OkCupid and its parent Match Group over data sharing from 2014. Clarifai also deleted AI models built with the data, confirming access to the images despite no admission of wrongdoing by OkCupid.