Today:

• Meta Expands into Smart Glasses

• Perplexity Launches Comet AI Browser

• OpenAI Poaches Tesla, Meta High-Ranking Engineers

• OpenAI to Launch AI Web Browser

• Anthropic Expands Claude for Education

Zuckerberg goes SCORCHED EARTH.... Meta BUYS Ray Bans for "AI Superintelligence"...

Mark Zuckerberg and Meta are aggressively building a Super Intelligence division, hiring top talent from OpenAI, Apple, Google DeepMind, Anthropic, and more. They tried to acquire Safe Superintelligence and Furiosa AI, but were rejected—showing how serious some startups are about staying independent. 

Meta also bought a stake in Ray-Ban’s parent company, signaling their plan to deliver AI through smart glasses. Expect major announcements at Meta Connect in September.

WATCH THE VIDEO ON YOUTUBE

Perplexity Launches Comet AI Browser

Perplexity has launched Comet, a new AI-powered web browser. Unlike traditional browsers, Comet lets users ask questions directly on any page, remembers context, and can complete tasks like booking meetings or comparing products. It’s designed to help people think, not just search. Making the internet feel like an extension of your own mind. Comet focuses on accurate answers and curiosity-driven use, aiming to turn online browsing into seamless, intelligent experiences.

Why This Matters

1. AI as Interface: Comet shifts the browser from a passive tool to an active thinking assistant, redefining how humans interact with the internet.
   

2. Agentic Execution: By handling full tasks (like booking or buying), Comet showcases the growing real-world power of autonomous AI agents.
   

3. Contextual Learning: Its personalized, memory-based reasoning hints at how future AI tools will adapt to individual thought patterns—boosting productivity and decision-making.

OpenAI Poaches Tesla, Meta High-Ranking Engineers

OpenAI has hired four top engineers from Tesla, xAI, and Meta, including Tesla’s former VP of software engineering, David Lau. These high-profile hires are joining OpenAI’s “scaling team,” as confirmed in a company Slack message by cofounder Greg Brockman. The move highlights OpenAI’s ongoing talent push to strengthen its infrastructure and maintain its lead in AI development, especially as competition intensifies with rivals like Elon Musk's xAI and Meta.

Why This Matters

1. Top Talent Migration: Poaching senior engineers weakens competitors while bolstering OpenAI’s technical depth in scaling AI systems.
   

2. Scaling Focus: Reinforces OpenAI’s priority to expand model capabilities and infrastructure, signaling the next leap in AI performance and deployment.
   

3. Competitive Pressures: Highlights the escalating talent war among AI giants, with staffing moves shaping the direction and speed of AI innovation.

OpenAI to Launch AI Web Browser

OpenAI is set to launch its own AI-powered web browser to compete with Google Chrome. Built on Chromium, it will feature chat and agent tools, keeping users in a ChatGPT-like interface to perform tasks like booking or searching. This gives OpenAI access to valuable user data, key to Google’s ad business. The move signals a deeper push into consumer tech, adding pressure in the AI race against Google and other rivals.

Why This Matters

1. Data Access Power: Owning a browser gives OpenAI direct access to user behavior data, critical for training and improving AI models.
   

2. Agent Integration: The browser will serve as a platform for AI agents, enabling real-time task execution like booking or shopping, marking a leap toward truly autonomous digital assistants.
   

3. Direct Challenge to Google: OpenAI entering the browser market hits Google at its core—ads, search traffic, and default control—potentially reshaping the internet’s balance of power.

🧠RESEARCH

SingLoRA: Low Rank Adaptation Using a Single Matrix

SingLoRA is a new method for fine-tuning AI models more efficiently. It uses just one compact matrix instead of two, avoiding instability and cutting the number of extra parameters in half. This leads to better results in tasks like reasoning and image generation, while using fewer resources than existing methods.

Agent KB: Leveraging Cross-Domain Experience for Agentic Problem Solving

Agent KB is a shared memory system that helps AI agents learn from each other. It stores both strategies and detailed actions, allowing agents to reuse knowledge across different tasks. This boosts performance significantly, with models like GPT-4 and Claude-3 solving harder problems more effectively by tapping into past experiences.

OmniPart: Part-Aware 3D Generation with Semantic Decoupling and Structural Cohesion

OmniPart is a new system for generating 3D objects with clearly defined, editable parts. Unlike traditional methods that create single, solid shapes, OmniPart separates objects into meaningful components while keeping them structurally sound. It allows user control over part design and delivers more flexible, accurate, and usable 3D models for interactive applications.

🛠️TOP TOOLS

Mindgrasp AI Free - Transform class lectures and readings into study tools with AI powered summaries, detailed notes, flashcards, quizzes, and an interactive AI-tutor trained on your course material.

ColorMagic - Generate color palettes with AI

HuxliAI - Designed to enhance and streamline academic work for students, leveraging advanced AI technologies to provide comprehensive support.

Amper Music AI - Generate a wide array of music types.

Conker AI - Designed to streamline the process of creating quizzes and assessments for teachers, particularly in the K-12 education sector.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Anthropic is expanding Claude for Education with new tools that let students access textbooks, lectures, and coursework directly in Claude. They’re also growing student programs and partnering with universities to promote responsible AI use.

• OpenAI is set to release a new open-weight AI language model—its first since GPT-2—allowing companies and governments to run it independently. The move may strain its exclusive partnership with Microsoft amid ongoing contract talks.

• Hugging Face launched Reachy Mini, a $299 open-source desktop robot aimed at making robotics affordable and accessible. Designed for developers, educators, and learners, it combines AI capabilities with community-driven hardware, challenging high-cost industry norms.

• OpenAI has finalized its $6.5 billion acquisition of Jony Ive’s hardware startup, io Products Inc., to develop AI-powered devices. Ive’s design firm LoveFrom will lead creative work, marking OpenAI’s serious push into consumer hardware.

• Chinese researchers unveiled MemOS, the first memory-focused operating system for AI, giving models human-like recall across conversations. It outperforms current memory systems, enables portable and evolving memory, and could redefine how AI learns, reasons, and adapts over time.

• Microsoft highlighted $500 million in annual savings from using AI tools across departments—even as it lays off thousands of employees. The company says AI boosts productivity, but the cuts raise questions about tech-driven job displacement.

• China plans to build 39 AI data centers in Xinjiang and Qinghai, aiming to use over 115,000 banned Nvidia chips despite U.S. export restrictions. While smuggling concerns rise, evidence of large-scale illicit procurement remains unclear U.S. officials are alarmed, as the effort underscores the limits of current chip export controls.