The 2,500 year-old Chinese game of Go, widely considered one of the most difficult and sophisticated board games in existence, was one of the last remaining games a computer had not yet mastered. The number of potential moves possible on the 19-by-19 grid exceeds the number of atoms in the entire universe. 

Two opponents -- black and white -- alternate play by placing their pieces, also called stones, on the grid. The object of the game is to capture the most spaces. Each player attempts to surround the pieces of his or her opponent to remove those pieces from the board and control the most territory. At the end of play, the opponent with the greatest number of spaces wins.

Three people sitting at a table while two play the Chinese game of Go.

The Chinese game of Go is one of the most simple yet complex games in the world. Image source: Getty Images.

DeepMind, a division of Alphabet's (NASDAQ:GOOGL) (NASDAQ:GOOG) Google, had its sights set on conquering the game, using its artificial-intelligence (AI) computer AlphaGo in the age-old battle of man vs. machine.

The end of an era

Last year, in a best-of-three series, AlphaGo deated 19 year-old prodigy Ke Jie, the world's top-ranked Go player. This was a battle of titans, and Google's own engineers reported: "According to AlphaGo's estimation of the match, the programme assessed the first 50 moves as virtually perfect, and the first 100 moves were the finest anyone has ever played against the Master version of AlphaGo." 

Google acquired AI start-up DeepMind in 2014 for an estimated $660 million, and this competition was the most recent in a series of human vs. AI matchups in the centuries-old game that began several years earlier. At the time, it was unthinkable that a computer could defeat a human competitor in such a profoundly complex game. Every previous attempt to design a computer system that could beat humans came up short.

That was before artificial intelligence.

AlphaGo consists of two AI-based deep neural networks, each a combination of software and algorithms that are modeled after the structure of the human brain and each with a specific task. The first was charged with narrowing down the nearly infinite number of moves to those that had the greatest likelihood of being played. The second analyzed those moves to determine which will most likely lead to a winning game. 

Deep Mind logo

Deep Mind isn't only about fun and games. Image source: DeepMind. 

Fun and games with real-world applications

DeepMind is currently focused more on AI research and less on business applications, but that isn't to say Google doesn't benefit financially from it. One unexpected example came to light last year, when the company revealed that by using deep learning, a specific discipline of AI, it was able to improve the energy efficiency of its data centers. Similar to how a laptop generates heat, data centers produce heat on a much larger scale that must be mitigated or reduced, producing enormous cooling costs.

By training the AI in the various operating scenarios, historical data, and average energy consumption, the system devised a method of alternating and optimizing the cooling system that achieved a 40% reduction in the amount of energy used for cooling. This achievement reduced the amount of energy the data centers used by a whopping 15%. That's significant when you consider that Google accounts for 0.01% of all global electricity use. 

Another example of the financial benefits of AI, also from last year, came when Google developed a specialized computer chip called the Tensor Processing Unit (TPU). This chip was able to outperform standard chips by being 30 to 80 times more efficient. These efficiency gains were so profound that it saved Google from having to build a dozen new data centers, nearly doubling its current footprint.

Google just released the latest version of its TPU, which will handle both the training of AI systems and running them once they're trained. It plans to make this technology available for free to researchers who share their results in peer-reviewed publications. Google also plans to make the upgraded technology available to customers through its Google Cloud.

Investors still win

It may sometimes be difficult for investors to wrap their minds around the ways in which a company like Google will financially benefit from AI. Keep in mind, however, that Google has a track record of using these algorithms to solve real-world problems, and while they may be difficult to quantify, investors stand to benefit nonetheless. We don't yet know what the next killer app will be, but Google and DeepMind believe that by focusing on the science, the financial gains will come -- and so far, they've been right.

Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool’s board of directors. Danny Vena owns shares of Alphabet (A shares). Danny Vena has the following options: long January 2018 $640 calls on Alphabet (C shares) and short January 2018 $650 calls on Alphabet (C shares). The Motley Fool owns shares of and recommends Alphabet (A shares) and Alphabet (C shares). The Motley Fool has a disclosure policy.