AI News Highlights for July 4th Week

Feb 12, 2025

AI News Update (Week of July 4th)

General Overview

  • Fewer major announcements due to the 4th of July holiday in the United States.
  • Still, several interesting updates and releases in AI.

Runway Gen 3

  • Gen 3 is now publicly available, but only for pro users.
  • Allows for text-to-video generation.
  • Example prompt used: “a bald eagle flying in front of an American flag with fireworks in the background.”
  • Comparison: Gen 3 is the best text-to-video generator currently, whereas Luma AI excels in image-to-video.

11 Labs Updates

  • New voices added to their reader app, including Judy Garland, James Dean, Bert Reynolds, and Sir Lawrence Olivier.
  • Voices added with permission and deals with respective estates.
  • Voice isolator feature introduced for improved audio quality.

Sunno App

  • Sunno app now available on iOS for easier music creation on mobile.
  • Works similarly to the web app.
  • Be cautious of fake apps claiming to be the official Sunno app.

Meta's 3D Text-to-Image Research

  • Research released on 3D Genin for generating 3D images from text prompts.
  • Potential applications in game development and 3D video asset creation.

Kotai's New Voice Model

  • Kotai released an open-source voice model aimed at competing with GPT-4’s voice capabilities.
  • Available for public testing at Moshi do chat.
  • Real-time responses but currently lacks expressiveness.

Open Source LLM: Intern LM 2.5

  • Released on Hugging Face with a 1 million context window.
  • Potential applications in custom model chat systems.

Brave Browser Update

  • Brave now allows bringing custom models into their AI, Leo AI.
  • Supported models include Mixr, Claude, and Llama.

Perplexity Pro Search Update

  • Multi-step reasoning and better math/programming due to Wolfram Alpha integration.
  • Usage limits: 5 times every 4 hours for free users and nearly unlimited for pro members.

Apple Joins OpenAI Board

  • Apple gains observer role without voting rights.
  • OpenAI facing another lawsuit from the Center for Investigative Reporting over copyright infringement.

Mustafa Solman’s Controversial Statement

  • Argument on fair use of content on the open web.
  • Cloudflare’s new feature to block AI scrapers.

Figma’s AI Features and Controversy

  • Figma to use user content to train AI models.
  • Controversy over similarity to Apple’s weather app.
  • Pause on the feature to avoid design replication issues.

YouTube and Instagram Updates

  • YouTube now allows takedown requests for AI-generated content simulating real people.
  • Instagram changes AI labeling to “AI info” to avoid confusion.

Grok 2

  • New version of Grok releasing in August with significant improvements.

Rumors of Apple and Google Gemini Partnership

  • Potential partnership to be announced in the fall.

WhatsApp New Feature Leak

  • Feature for generating cartoon or alternate versions of user images.
  • Likely to roll out to other Meta platforms.

Meta Ray-Ban Competitor

  • New company developing similar glasses with ChatGPT-4 integration.

Open Telekinesis Demo at MIT

  • Remote operation of robots demonstrated using Apple Vision Pro.
  • Example: operating a robot in UCSD from MIT.

Closing Remarks

  • Encouragement to follow Future Tools for latest AI news and tools.
  • Future newsletter to cover broader emerging technologies.
  • Gratitude for the audience and call to subscribe for more updates.