How does Alpha Zero compare to Deep Blue ?
Why does it matter, especially for teenagers?
Deep Blue was a specialized, powerful chess calculator, while Alpha Zero is a versatile, self-learning game player that can master multiple games on its own.
Deep Blue (1997):
Domain: Chess
Approach: Deep Blue was like a super-smart calculator. It evaluated positions on the chessboard using pre-programmed rules and databases of known moves.
It had a brute-force strategy, exploring millions of positions per second to find the best move.
The opening book encapsulated more than 4,000 positions and 700,000 grandmaster games, while the endgame database contained many six-piece endgames and all five and fewer piece endgames. An additional database named the “extended book” summarizes entire games played by Grandmasters. The system combines its searching ability of 200 million chess positions per second with summary information in the extended book to select opening moves….The system derived its playing strength mainly from brute force computing power…
Domains: Chess, Shogi, Go, and more
Approach: Alpha Zero is more like a learning brain. It doesn’t start with pre-programmed strategies. Instead, it learns by playing against itself.
Imagine a child learning a game without any instructions. Alpha Zero uses deep neural networks to understand the game and improve through trial and error.
After four hours of training, DeepMind estimated Alpha Zero was playing chess at a higher Elo rating than Stockfish 8; after nine hours of training, the algorithm defeated Stockfish 8 in a time-controlled 100-game tournament…
Wiki
Comparison:
Adaptability: Deep Blue is designed for chess only, while Alpha Zero can tackle various games, learning and mastering them without human guidance.
Learning: Deep Blue relies on human-programmed knowledge, whereas Alpha Zero teaches itself, making it more flexible and capable of learning new games.
How it matters to you ?
In the late 1990s, personal computers became a common presence in households. By 1997, a computer successfully defeated Kasparov in chess, and by 2017, we witnessed the emergence of Alpha Zero.
Out of the 5 human senses — Machines today can see and hear. At this rate very soon they would be able to taste, touch and taste. They have immense processing and memory at their disposal.
One missing bit is Human Values — but here also a considerable work is under construction. Example, check this book ; The Alignment Problem: How Can Machines Learn Human Values? by Brian Christian (Author)
I took these two examples to illustrate how fast the technology evolves in a lifetime. Within roughly three decades, machines are sitting on our tables and have progressed to the point of autonomous learning. If you’re a teenager about to finish high school, it’s crucial to closely monitor the AI revolution, considering its potential impact on your career over the next five decades of your life.
This is particularly important because technology tends to grow exponentially, while human thinking often follows a linear trajectory.
🍁Hope you found this useful.
