AI News Lecture Notes

Introduction

• Weekly updates on practical AI use cases.
• New developments in AI technology.

Google's Gemini 2.5 Pro

• New model capable of turning screen recordings into applications.
• Improved front-end capabilities.
• Can rebuild applications from video recordings.

Testing Gemini 2.5 Pro

• Attempted recreation of a time converter app using video context.
• Initial results were promising but required follow-up prompts.
• Challenges with perfect replication, but iterative improvements possible.

ChatGPT Updates

• Developer-focused improvements.
• New GitHub connection for deep research.
• Updated guides for model usage.

AI Advantage Community

• Offers courses and resources on AI use cases.
• Includes a fine-tuning course for customizing models.
• Community for support and learning.

MidJourney's Omni Reference

• New feature for product photography.
• Allows referencing of an image across multiple creations.

Nvidia's Parakeet

• Open-source transcription model for English.
• Offers fast and accurate transcription with timestamps.

Hen's AI Video Avatar

• Can create AI avatars from a single image.
• Simplifies model training for personal avatars.

Suno 4.5 Music Creation

• New AI model for generating high-quality music.
• Allows creation of longer songs with better prompt adherence.

Quickfire AI News

• Notebook announces desktop app for long context consumer use.
• OpenAI acquires Windsurf, enhancing AI-powered code editing.
• LTX releases open-source video models.
• Community-created philosophical space shooter game.

Visa and Mastercard Agentic Technology

• Integrating agentic payment technology for AI commerce.

Conclusion

• Updates on AI tools and their applications.
• Resources available via newsletter signup.