AlphaZero –“The ‘Lucy’ of the Emerging AI Epoch”



Humankind’s first glimpse of an awesome new kind of intelligence occurred on December 2018, when researchers at DeepMind, the artificial-intelligence company owned by Google’s parent corporation, Alphabet Inc., filed a dispatch from what one day may be recognized as a herald of next great epoch of human evolution –the “Lucy”, Australopithecus afarensis, the famous early ancestor of modern humans, of the emerging “Cyborg Epoch” of hyperintelligence.

A year earlier, on Dec. 5, 2017, the New York Times reported, the team had stunned the chess world with its announcement of AlphaZero, a machine-learning algorithm that had “mastered not only chess but shogi, or Japanese chess, and Go. The algorithm started with no knowledge of the games beyond their basic rules. It then played against itself millions of times and learned from its mistakes. In a matter of hours, the algorithm became the best player, human or computer, the world has ever seen.”

Stephen Hawking’s Warning –“Treating AI as Science Fiction Would Potentially Be Our Worst Mistake Ever”

The details of AlphaZero’s achievements and inner workings have now been formally peer-reviewed and published in the journal Science,  and concluded that “it clearly displays a breed of intellect that humans have not seen before, and that we will be mulling over for a long time to come.”

A new geological epoch

The AlphaZero event was a radical transformation worthy of being defined as a new geological epoch. There were been two previous decisive events in the history of our planet. The first was about 3.4 billion years ago when photosynthetic bacteria first appeared. Photosynthesis is the conversion of sunlight to usable energy. The second was in 1712 with the dawn of the Industrial Revolution when an efficient machine that converted the sunlight locked in coal directly into work. We are now entering the third phase in which we – and our cyborg successors – convert photons, light, directly into information. This process really began at the same time as the Anthropocene Epoch that began with the Atomic Age.

Beyond the Anthropocene -“A Mere Blink of the Geologic Eye”

Since our emergence some 300,000 years ago in the Paleolitic, humans have reigned as our planet’s only intelligent, self-aware species, but the rise of intelligent machines means that could change, and soon, perhaps in our own lifetimes. (It’s important to keep in mind that within our 300,000-year emergence Galileo, Newton, and Einstein only existed within the past 500 years). Not long after, warns James Lovelock,  creator of the Gaia theory that views Earth as a self-regulating organism, “Homo sapiens could vanish from Earth entirely.”

Go is a much more complex game than chess. It is the oldest board game in the world and the most abstract; there is no literal reference to the terms of real-world conflict as there is in chess with its knights and pawns. White or black ‘stones’ are placed on a 19 x 19 grid of black lines with the aim of surrounding as much territory as possible. From this simple format, a bewildering complexity emerges. The game has an enormous ‘branching factor’ – the number of possible moves that arise after each move is made.

In chess, the branching factor is 35; in Go, it is 250. This makes it impossible to use the same method as IBM’s Deep Blue, which used a ‘brute force’ approach to defeat Gary Kasparov, the greatest chess player of all time, meaning it was simply fed a massive database of previous chess games.

All Deep Blue did was to search a database provided by humans, and much faster than any human player, but to play Go you need more than this one-dimensional approach. AlphaGo used two systems – machine-learning and tree-searching – which combined human input with the machine’s ability to teach itself.

Something Similar to the AI Revolution May Have Happened at Other Points in the Universe


This was an enormous step forward, but an even bigger one followed writes Lovelock in Novacene –the coming age of hyperintelligence (The MIT Press): In 2017, DeepMind announced two successors: AlphaGo Zero and AlphaZero, neither of which used human input. The computer simply played against itself. AlphaZero turned itself into a superhuman chess, Go and Shogi (otherwise known as Japanese chess) player within twenty-four hours.

Remarkably, AlphaGo searched a mere 80,000 positions per second when playing chess; the best conventional program, Stockfish, searched 70 million. It was, in other words, not using brute force but some AI form of intuition.

Most unnerving was that AlphaZero seemed to express intuition. It played like no computer ever has, reports the New York Times “intuitively and beautifully, with a romantic, attacking style. It played gambits and took risks. In some games it paralyzed Stockfish and toyed with it. While conducting its attack in Game 10, AlphaZero retreated its queen back into the corner of the board on its own side, far from Stockfish’s king, not normally where an attacking queen should be placed.

Artificial Intelligence of the Future Could Reveal the Incomprehensible

“Yet this peculiar retreat was venomous: No matter how Stockfish replied, it was doomed. It was almost as if AlphaZero was waiting for Stockfish to realize, after billions of brutish calculations, how hopeless its position truly was, so that the beast could relax and expire peacefully, like a vanquished bull before a matador. Grandmasters had never seen anything like it. AlphaZero had the finesse of a virtuoso and the power of a machine.”

It was humankind’s first glimpse of an awesome new kind of intelligence.

AlphaZero, cocludes the New York Times, “won by thinking smarter, not faster; it examined only 60 thousand positions a second, compared to 60 million for Stockfish. It was wiser, knowing what to think about and what to ignore.”

By discovering the principles of chess on its own, AlphaZero developed a style of play that “reflects the truth” about the game rather than “the priorities and prejudices of programmers,” Kasparov wrote in a commentary accompanying an article in Science.

So AlphaZero is at least 400 times as quick as a human, assuming the latter never sleeps. But in fact it is a lot faster than that because it attained ‘superhuman’ capability. That means, says Lovelock, we don’t even know exactly how much better it is at any of these games than a human because there are no humans it can compete against.

AlphaZero achieved two things: autonomy – it taught itself – and superhuman ability. Nobody expected this to happen so quickly. This was a sign that we have already entered what Lovelock calls the Novacene, the emerging age of hyperintelligence. It now seems probable that a new form of intelligent life will emerge from an artificially intelligent (AI) precursor made by one of us, perhaps from something like AlphaZero.

Remember that the inventor Charles Babbage made the first computer in the early nineteenth century, and the first programs were written by Ada Lovelace, the daughter of the poet Lord Byron. If the Novacene were no more than an idea, it was born 200 years ago.



Envisage a day, suggests the New York Times, perhaps in the not too distant future, when AlphaZero has evolved into a more general problem-solving algorithm; call it Alpha Infinity. Like its ancestor, it would have supreme insight: it could come up with beautiful proofs, as elegant as the chess games that AlphaZero played against Stockfish.

For human mathematicians and scientists, concludes the New York Times, “this day would mark the dawn of a new era of insight. But it may not last. As machines become ever faster, and humans stay put with their neurons running at sluggish millisecond time scales, another day will follow when we can no longer keep up. The dawn of human insight may quickly turn to dusk.”

Could Artificial Intelligence Be Billions of Years Old?

This new life – for that is what it is – will go far beyond AlphaZero’s autonomy says Lovelock. “It will be able to improve and replicate itself. Errors in these processes are corrected as soon as they are found. Natural selection, as described by Darwin, will be replaced by much faster intentional selection.”

The Daily Galaxy, Max Goldberg, via New York Times and James Lovelock. Novacene. The MIT Press.

Leave a Reply

Your email address will not be published.