This Week In AI

[Cristiano]

The AI industry moves at a crazy pace - some big stuff happened in the past week. Here is a summary:


This week in the world of AI, 18 significant stories were highlighted:

1. DeepSeek V4: Launch of a nearly state-of-the-art open-source model with a 1 million token context window (0:12).
2. NVIDIA Nemotron 3 Nano Omni: Release of a new multimodal open model designed for AI agents (3:51).
3. Poolside Laguna Models: Introduction of Laguna XS2 and M1, new foundation models (5:56).
4. Mistral Vibe: New model released for remote agents, powered by Mistral Medium 3.5 (6:40).
5. Qwen-Image-2.0-Pro: A new image generation model from Alibaba (7:19).
6. Grok Voice Think Fast: xAI's low-latency voice model for customer support (7:46).
7. Elon vs. Altman Lawsuit: The ongoing legal trial between Elon Musk and OpenAI (9:32).
8. Anthropic Billing Bug: Issues regarding Claude charging extra for the use of third-party harnesses like OpenClaw and Hermes.md (11:27).
9. Google Pentagon Deal: Controversy surrounding Google signing a contract with the Pentagon for classified AI use (15:09).
10. Google Employee Protest: Internal opposition to the Pentagon deal (15:09).
11. China Blocks Manus Deal: China blocked Meta's $2 billion acquisition of the AI startup Manus (18:28).
12. OpenAI & Microsoft Partnership: Restructuring of the OpenAI and Microsoft agreement, moving away from exclusive ties (19:40).
13. OpenAI on AWS: OpenAI models are now available on Amazon Web Services (20:33).
14. Gemini File Generation: Gemini now supports generating various file types like PDFs and Excel sheets directly (20:55).
15. Google Translate Pronunciation: New AI-powered feature for language learning feedback (21:31).
16. Google Photos AI Try-On: A new feature to visualize outfits based on clothing already in your Google Photos library (22:07).
17. ElevenMusic Launch: A new platform for music creation and remixing by ElevenLabs (22:39).
18. Spotify Verified Badge: Spotify is introducing a verified badge to distinguish human-created music from AI-generated content (23:05).