🤖 Google DeepMind Is Paying AI Staff to Do Nothing

Subscribe | AI Toolkit | Meet The Team

Welcome back AI enthusiasts!

In today’s Daily Report:

• 🧯Google DeepMind Is Paying AI Staff to Do Nothing

• 🦾LLMs Solve Complex Planning Challenges

• 🛠Trending Tools

• 🥪Brief Bites

• 💰Funding Frontlines

• 💼Who’s Hiring?

Read Time: 3 minutes

🗞RECENT NEWS

AI TALENT WARS

🧯Google DeepMind Is Paying AI Staff to Do Nothing

Google DeepMind is allegedly paying some AI Staff to do nothing for a year rather than join rivals like OpenAI and Anthropic.

Key Details:

• As AI Talent becomes more valuable, salaries are skyrocketing. For instance, top-tier AI Researchers command salaries exceeding $500,000 annually, and AI Engineers with a few years of experience earn between $150,000 and $300,000 annually.

• With the AI Job Market becoming more competitive than ever, Big Tech companies are adopting strategic hiring practices and implementing retention methods to secure AI Talent.

• Google DeepMind has reportedly resorted to using “aggressive” non-compete agreements for some AI Staff in London, UK. So, why do this in Europe and not America? Certain States within the U.S. have a complete ban on non-compete agreements, including California.

• These “aggressive” non-compete agreements bar the AI Staff from working for rivals for up to a year after leaving Google DeepMind, even paying them a full salary during that period.

Why It’s Important:

• This strategy aims to prevent AI Staff who leave Google DeepMind with valuable knowledge and skills from immediately joining rivals. As AI developments continue, the competition for AI Talent will only intensify.

• Microsoft’s VP of AI posted on X about how Google DeepMind’s AI Staff have been reaching out to him in “despair,” trying to escape their “aggressive” non-compete agreements.

🩺 PULSE CHECK

AI RESEARCH

🦾LLMs Solve Complex Planning Challenges

Researchers at the MIT-IBM Watson AI Lab developed “LLM-Based Formal Planning (LLMFP),” which helps LLMs solve complex planning problems.

Key Details:

• Imagine a small bakery is trying to create a weekly schedule for baking five different types of pastries. They’ve got limited oven space, different baking times for each pastry, varying ingredient costs for each pastry, and fluctuating demand throughout the week for each pastry.

• The goal is to ensure they don’t exceed oven space at any given time throughout the week and meet the minimum daily demand for each pastry.

• The small bakery tried leveraging OpenAI’s ChatGPT, which is powered by GPT-4, to help them. Unfortunately, LLMs like GPT-4 struggle to solve this type of complex planning problem.

• Here’s a concise breakdown of how LLMFP works:

  1. The Input: You provide a brief text description of the complex planning problem, including entities, resources, actions, constraints, and objectives.

  2. Structure and Refine: LLMFP formalizes the brief text description, identifying decision variables, outlining key constraints, and creating manageable sub-problems.

  3. Integrate and Validate: The LLM solves the manageable sub-problems and combines each solution into a complete answer.

Why It’s Important:

• LLMFP achieved an 83.7% success rate when helping GPT-4o (“o” for “omni”) solve complex planning problems, boasting a 37.6% improvement.

• It essentially supercharges LLMs to act as smart virtual assistants that excel at solving complex planning problems. “It can figure out the best plan that meets all your needs,” explained Yilun Hao, a Graduate Student at the MIT Laboratory for Information and Decision Systems (LIDS).

🛠TRENDING TOOLS

🪄Readdy builds your dream website.

🧠TwinMind is your AI-powered digital twin brain.

🎬kozy.studio generates high-quality edits in seconds.

🔍iAsk is an AI-powered search engine that delivers true answers.

⚙️GitSummarize turns any GitHub repository into internal documentation.

🔮Browse our always Up-To-Date AI Tools Database.

🥪BRIEF BITES

OpenAI has reportedly explored acquiring Legendary Apple Designer Jony Ive’s AI Hardware Startup for $500 million.

Runway introduced “Gen-4 Turbo,” which turns text prompts or static images into 10-second videos within 30 seconds.

Shopify CEO Tobias Lütke told employees they must prove their jobs can’t be streamlined with AI Resources before asking for more headcount.

Former OpenAI CTO Mira Murati’s new AI Lab, Thinking Machines, gained Ex-OpenAI Chief Research Officer (CRO) Bob McGrew as an advisor.

💰FUNDING FRONTLINES

• Ropes closes a $3.1M Seed Round for scaling AI-Driven Skills-Based Hiring.

• Qevlar AI raises a $14M Series A to Supercharge Security Operations with Agentic AI.

• SchoolAI secures a $25M Series A for an AI-Enabled Learning Platform that provides personalized learning for every student.

💼WHO’S HIRING?

• NVIDIA (Santa Clara, CA): Deep Learning (DL) Compiler Intern, Summer 2025

• Point72 (New York, NY): L/S Data Analyst, Entry-Level

• Distyl AI (San Francisco, CA): AI Engineer, Mid-Level

• Anthropic (San Francisco, CA): Safeguards Policy Analyst, Senior-Level

📒FINAL NOTE

FEEDBACK

❤️TAIP Review of The Day

REFER & EARN

🎉Your Friends Learn, You Earn!

You currently have 0 referrals, only 1 away from receiving ⚙️Ultimate Prompt Engineering Guide.

Copy and paste this link to friends: https://theaipulse.beehiiv.com/subscribe?ref=PLACEHOLDER