A study of mental health conversations with ChatGPT, Claude, Gemini, and Meta AI: they often failed to recognize signs of conditions and offered general advice (Georgia Wells/Wall Street Journal)

As artificial intelligence tools like ChatGPT gain popularity for personal therapy and emotional support, concerns have arisen regarding their use, particularly among young people. Employers are increasingly utilizing generative AI to evaluate employees' psychological well-being and provide emotional support in the workplace, raising questions about the implications for privacy and effectiveness.

The Allen Institute for AI (Ai2) has launched its latest large language models, Olmo 3, claiming they outperform existing open models such as Stanford's Marin and commercial models like Llama 3.1. This new generation aims to enhance performance and efficiency in AI applications.

Gemini, Google's AI model, has been integrated into Android Auto, enhancing the platform's capabilities. This integration allows for improved user interactions and functionality within vehicles, marking a significant step in the evolution of automotive technology.

Gemini, Google's new AI model, is being rolled out globally to Android Auto, replacing Google Assistant. This integration allows drivers to create playlists, access emails, and learn about their surroundings using voice commands.

Google has made advanced features of its Gemini AI model available for free to all Gmail users through Google Vids. Previously, these features were exclusive to paid subscribers, enhancing accessibility for a wider audience.

Large language models (LLMs) like ChatGPT have improved significantly in their ability to handle complex reasoning and math problems, which were previously challenging for them. This advancement marks a notable shift in their functionality, enhancing their utility in various applications.

Google has announced that its Gemini app can now detect images created or edited by its AI technologies. The company also plans to introduce verification features for video and audio content in the near future. This development aims to enhance the app's capabilities in identifying AI-generated media.

OpenAI has released the Agentic Codex model, which is capable of functioning for over 24 hours. This new model is available to users on ChatGPT Plus, enhancing the capabilities of AI-assisted coding and software engineering tasks. The introduction of this model marks a significant advancement in OpenAI's efforts to improve coding efficiency and problem-solving abilities.