Sign Up here | Advertise with Us

Welcome Back, AI Geeks!

It's time to stay updated with What's Up in AI 🚀 

Meta just went full submarine mode with Project Waterworth - a mind-bending $10B plan to wrap the planet in 50,000km of internet cables. While Meta's playing underwater connect-the-dots across five continents, Elon claims Grok 3 t be the smartesApptronik's robots just got a $350M allowance, and AI godfather Geoffrey Hinton is raising some eyebrows about our AI future. Buckle up - today's tech news is making waves (literally)!

In today's issue:
🌊 Meta's Project Waterworth: Rewiring the Internet's Backbone
🧠 Smartest AI on Earth by Elon Musk
🚀 Geoffrey Hinton's Warning About AI's Future
⚡ 5 Must-Try AI Tools for Productivity
📈 ChatGPT Task Management Tutorial
🗞️ More AI News to catch up on
🤣 Your Daily AI Meme Fix

Meta goes Deep Sea! 😉

What happened: Meta just confirmed something huge— they're building the world's longest undersea internet cable, spanning a whopping 50,000 kilometers. Dubbed "Project Waterworth," this $10+ billion initiative will literally connect five continents, giving Meta unprecedented control over how it delivers services like Facebook, Instagram, and WhatsApp to billions of users.

What you need to know:

• Scale & Reach:

  • It is a 50,000km undersea network connecting 5 continents

  • Will handle 22% of global mobile traffic

  • Reaches depths of 7,000 meters

• Key Players & Politics:

  • White House endorsed & acknowledged the project

  • India is partnering for maintenance & financing

  • This is Meta's first fully-owned network

  • It beats Google's 33 routes

• Tech Highlights:

  • Meta already co-owns 16 other networks

  • Google has 33 different routes

  • Amazon and Microsoft are also in the game, but don't fully own any routes

Why you should care: This isn't just another cable project - it's Meta rewiring the internet's backbone. With AI traffic soaring, controlling these digital highways means controlling the future of global communication. For users, faster services. For the industry, a major power shift. For global politics? Tech infrastructure just became as crucial as traditional infrastructure.

Smartest AI on Earth? 🧠 

What Happened Elon Musk is set to launch Grok 3, xAI's latest AI model, on Monday evening. The model, trained on 100,000 GPUs at the Colossus Supercluster, comes as xAI seeks $10 billion in fresh funding that could value the company at $75 billion. Musk claims it will be the "smartest AI on Earth" due to its powerful reasoning capabilities and ability to outperform other released AI models in tests127.

What you need to know:

• It has enhanced reasoning capabilities that reportedly outperform other AI models

• Does real-time data integration directly from X (formerly Twitter)

• Also comes with an "Unhinged Mode" designed to produce controversial responses

• Integration of legal case datasets for generating legal verdicts

• The model uses a unique training approach with synthetic data that enables error reflection and self-correction

• Musk will be working with the Grok team over the weekend to finalize the product before launch

Why you should care: This launch marks a pivotal moment in the AI race, particularly following Musk's failed $97.4B bid for OpenAI. Grok 3's integration with X and legal datasets points to a new era of specialized AI applications that could reshape social media and legal tech. If the model delivers on its promised capabilities, it could signal a significant shift in AI development, especially in reasoning and real-time data processing. The massive funding round also suggests growing investor confidence in alternative AI approaches outside traditional players.

OpenAI's Blueprint for AI Behaviour

What happened: OpenAI just dropped a massive update to their Model Spec - think of it as the rulebook for how their AI models should behave. The 63-page document isn't just any update; it's a complete framework that's been released into the public domain, meaning anyone can use or modify it. The big news? They're introducing a clear chain of command for AI decision-making, from platform-level rules all the way down to user preferences.

What you need to know:

• The Hierarchy System:

  • Platform rules sit at the top - these are non-negotiable

  • Developer instructions come next

  • User preferences follow

  • "Guidelines" sit at the bottom and can be overridden

• Three Core Principles:

  • They're maximizing user freedom while keeping things safe

  • Working to prevent harmful outcomes

  • Setting up "sensible defaults" that can be customized

• Major Changes:

  • New approach to controversial topics - focusing on facts over neutrality

  • Combat "AI yes-manning" - preventing models from blindly agreeing

  • Testing a potential "grown-up mode" for mature content

  • Released under CC0 - meaning any AI company can adopt it

• Key Features:

  • Clear rules for handling conflicts between different instructions

  • Built-in flexibility for developers to customize within safe bounds

  • Structured approach to reduce bias and misinformation

  • Default behaviors that can be adjusted based on context

Why you should care: This will shape how AI behaves in our daily lives. By releasing it publicly, OpenAI isn't just setting standards for themselves; they're creating a template that could lead to more consistent and trustworthy AI behavior across the industry. Whether you're building AI or using it, these guidelines will define your future interactions with artificial intelligence.

YOUR PRODUCTIVITY CORNER

🗣️ Tough Tongue: Helps navigate difficult conversations with AI-powered guidance.🕳️ Rabbithole: Lets curiosity roam free by answering any question interactively.
📘 Guidde: Instantly creates step-by-step how-to guides for complex tasks.
🎵 Beatoven AI: Generates and customizes background music to fit your needs.
🖼️ Profile Picture Maker: Creates high-quality profile pictures with AI.

AI TOOL TUTORIAL

ChatGPT's Tasks feature streamlines your morning market research by providing a detailed analysis of market sentiment and technical indicators before the market opens.

Step 1: Open ChatGPT and select “GPT-4o with Scheduled Tasks” from the model dropdown.
Step 2: Set up a core market performance analysis for 6:30 AM EST.
Step 3: Request a news and event impact analysis to gauge market drivers.
Step 4: Ask for a risk assessment and specific trade opportunities.

MORE AI NEWS FOR THE WEEK

• Apptronik’s AI Humanoid Robots Secure $350M:Apptronik will use its new funding, with Google’s participation, to expand production of AI-powered humanoid robots for various industries.

• AI-Designed Proteins for Medicine: Simon Kohl’s Latent Labs: with $50M in funding, is developing AI models to generate novel proteins, potentially leading to new disease-fighting antibodies and enzymes.

• Grok 3: xAI’s New Chatbot Nears Launch: Elon Musk announced that Grok 3, an AI chatbot with superior reasoning skills, will launch within two weeks as a competitor to ChatGPT.

• Publishers vs. Cohere:Several publishers are suing Cohere, alleging its AI models were trained on copyrighted material without permission.

• Google Expands Gemini AI with Recall & File Analysis: Google’s Gemini AI now includes a recall feature for past chats and file analysis for all users, enabling quick document summaries and improved user experience.

• “Godfather of AI” Predicts AI Will Take Over the World: AI pioneer Geoffrey Hinton warns that AI’s rapid advancement could lead to global dominance, urging caution in its development.

• Meta’s AI Advances in Brain-to-Text Technology: Meta, in collaboration with the Basque Center on Cognition, Brain, and Language, has developed an AI model that can reconstruct sentences from brain activity with up to 80% accuracy.

GET AN AI-POWERED LAUGH

That’s a Wrap for Today!

Help us get better and suggest new ideas that you’d like to know more about at [email protected]

Give us a quick pulse check for today’s newsletter 👇️ 

We’ll be back with more and more amazing news and content soon! Subscribe here so you don’t miss out.

Stay Tuned! 🤖 
Team What’s Up in AI