OpenAI and Broadcom announced Jalapeño, an AI accelerator designed to improve performance per watt for large language models, with early testing showing significant gains over current state-of-the-art solutions.
Princeton researchers use AI to design radio-frequency integrated circuits, achieving record performance and reducing design time by orders of magnitude.
Pangram CEO Max Spero claims language models expose themselves by repeating similar arguments, a key flaw in AI-generated text detection.
Bright Data, a web data collection platform, says the data suggests there's far more data out there than previously thought, with 97% of AI organizations relying on real-time web data infrastructure.
Qualcomm has acquired Modular, a chip startup, for nearly $4 billion, including $300 million for employee compensation.
Hugging Face introduces DoctoBERT, a French medical encoder, trained on a new corpus of web data curated with domain-specific filters and rephrasing techniques.
Mistral AI's new OCR 4 model outperformed competitors in 72% of blind test cases, according to the company. The model supports 170 languages and is available via API, Studio, and Microsoft Foundry.
Anthropic's Claude Tag, now in beta, writes 65% of internal code for its product team, the company says.
HuggingFace's Moon Bot streamlines engineering workflows by integrating Slack with internal tools, enabling real-time data access and persistent memory storage.
Oracle reported 21,000 job cuts in a year, citing AI adoption and debt-fueled expansion as key factors.
Anthropic launched Claude Tag on June 23, 2026, allowing teams to delegate tasks to the AI model within Slack channels, with 65% of internal code now created by an internal version of the tool.
