DeepMind's David Silver Secures $1.1 Billion for AI Superlearner Independent of Human Data
David Silver, formerly of DeepMind, raised $1.1 billion to develop a superlearner AI that acquires intelligence without relying on human-generated data. The project aims to create a foundational 'law of intelligence' akin to Darwin's law of natural selection. This approach promises to generate its own training data through self-directed exploration, as stated on the company's website.
This breakthrough shifts AI development from data dependency to autonomous learning paradigms. You will rethink workflows by prioritizing self-supervised systems over curated datasets, reducing costs and biases from human data. It teaches the principle of intrinsic motivation in AI, enabling scalable intelligence without endless data scraping.
David Silver at his new venture (ex-DeepMind) raised $1.1B in funding from top VCs including a16z and Sequoia, positioning it as a contender to redefine AI foundations with zero-shot learning capabilities.
Step 1: Install Curiosity-driven exploration library like IC3Net via pip install ic3net (https://github.com/pathak22/noreward-rl). Step 2: Train a simple agent on a gridworld environment using Python script: define intrinsic reward as prediction error of next state. Step 3: Run simulation for 1M steps; expect agent to discover novel paths 5x faster than reward-only baselines, mimicking data-free learning.