Luma AI releases Ray2 generative video model with ‘fast, natural’ motion and better physics

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

1. Luma AI releases Ray2 generative video model with ‘fast, natural’ motion and better physics

Luma AI, a San Francisco-based company, has released its latest video AI generation model, Ray2, which offers "fast, natural coherent motion and physics" and was trained with 10 times more compute than the original Luma AI video model, Ray1.

The model is available through the Dream Machine website and mobile apps for paying subscribers. The Dream Machine web platform offers a free tier with 720 pixel generations capped at a variable number each month, while paid plans begin at $6.99 per month. The model can generate new videos in a matter of seconds, although it can take minutes at a time due to a crush of demand from new users

Read the full story here

2. Humanoid robot maker Neura robotics raises US$120 million round

German startup NEURA Robotics has raised $120 million to improve its robots' perception and processing abilities. The funding was led by Lingotto Investment Management and other investors.

The company is expanding, doubling its workforce to 300 in the last year. The company is targeting the market, which is dominated by Asian and American companies, with nearly $1 billion in orders. NEURA Robotics' robots are expected to save on labor costs and shortages. The market is still in its pilot stage.

Read the full story here

3. Up Network and DreamSmart partner on Web3 AI glasses powered by Google Gemini

Up Network has partnered with DreamSmart to create Web3 AI glasses powered by Google Gemini. These glasses, designed under DreamSmart's StarV brand, offer a seamless, intuitive experience with real-time context-aware intelligence.

They weigh 44 grams and provide up to eight hours of battery life. The glasses will be available in Q1 2025 and will be powered by AI, XR capabilities, and Web3 incentives. The company claims they surpass current offerings from major tech players like Google and Samsung.

Read the full story here

4. AutoGen v0.4: Reimagining the foundation of agentic AI for scale, extensibility, and robustness

Microsoft has announced AutoGen v0.4, a complete redesign of the AutoGen library to improve code quality, robustness, generality, and scalability in agentic workflows. The update addresses architectural constraints, inefficient APIs, and limited debugging and intervention functionality.

It includes features like asynchronous messaging, modularity, observability and debugging, scalable and distributed agent networks, built-in and community extensions, cross-language support, and full type support. The framework supports both first-party and third-party applications and extensions.

Read the full story here

5. OpenAI announces ChatGPT Tasks for automating future actions

OpenAI has introduced a new beta feature called Tasks, transforming ChatGPT into a proactive AI assistant capable of managing scheduled tasks. Users can instruct ChatGPT to perform future actions or provide reminders at scheduled times, offering a more personalized and timely user experience.

Tasks can be scheduled directly within chat or through the "Tasks" section in the profile menu. The feature is rolling out globally to premium customers, starting with web access.

Read the full story here