- Alphabet's Go program AlphaGo defeated a top rated human champion.
- The deep learning AI technologies showcased in Alphabet's AlphaGo have important applications and could displace skilled human workers.
- Products and services powered by AlphaGo technology could begin to appear in one to two years.
Alphabet Inc-A (NASDAQ:GOOGL) claimed a major Artificial Intelligence (AI) breakthrough, The Wall Street Journal reports. Alphabet unveiled a software program that taught itself to beat a top human player of the board game Go, considered a milestone AI challenge.
AI programs for chess have beaten top human players since 1997, but Go, a Japanese board game in which players place stones on a square 19-by-19-line grid, is a bigger challenge because there are far more possible moves. The rules are simple - players take turns to place black or white stones on a board, trying to capture the opponent's stones or surround empty space to make points of territory - but Go is much more complex than chess, and experts predicted it would be at least another 10 years until a computer could beat one of the world’s elite group of Go professionals. Until last week.
Alphabet's program AlphaGo defeated a top Go champion. Alphabet's breakthrough is detailed in a research article titled "Mastering the game of Go with deep neural networks and tree search," published in the prestigious Nature journal.
AlphaGo, a Go program developed by Alphabet's company Google DeepMind, uses deep neural networks to mimic expert players, and further improves its performance by learning from games played against itself. AlphaGo has achieved a 99 percent win rate against the strongest Go programs, and defeated the reigning European champion Fan Hui 5–0 in a tournament match.
Google DeepMind is a British AI company founded in 2010 as DeepMind Technologies and acquired by Google in 2014. Google DeepMind has created neural networks that learn how to play games in a similar fashion to humans and appears to mimic key cognitive aspects of the human brain.
The Nature paper "describes a new approach to computer Go that combines Monte-Carlo tree search with deep neural networks that have been trained by supervised learning, from human expert games, and by reinforcement learning from games of self-play," notes the DeepMind website. "This is the first time ever that a computer program has defeated a human professional player. In March 2016, AlphaGo will face its ultimate challenge: a 5-game challenge match in Seoul against the legendary Lee Sedol, the top Go player in the world over the past decade."
Top Silicon Valley tech giants are spending billions on neural networks, machine learning, and other AI endeavors. IBM (NYSE:IBM), the company that developed the first AI program able to defeat top human players at chess in 1997 and Jeopardy! in 2011, recently expanded its AI platform Watson with new AI tools in the Watson Developer Cloud. Facebook (NASDAQ:FB) is becoming one of the most advanced AI research centers in the world, and Facebook's AI developments have accelerated to the point that AI will surely shape the way humans interact with computers.
Wired reports that Facebook places enormous importance on the sort deep learning AI that powers Alphabet's AlphaGo, and notes that Alphabet and Facebook are intent on building an artificial intelligence that will, in many ways, exceed the intelligence of humans.
AI experts find Alphabet's AlphaGo a significant breakthrough, Of course, most experts caution that AlphaGo is still far from a high-level "Artificial General Intelligence" (AGI) program that can train to solve virtually any sort of problem, but progress is expected. Google DeepMind co-founder and CEO Demis Hassabis said that products and services powered by AlphaGo technology could begin to appear in one to two years and include sophisticated smartphone-based intelligent personal assistants, and software that helps doctors diagnose illnesses.
The impact of learning AGI systems is likely to be very disruptive. In fact, experts warn that programs capable of humanlike combinations of insight, pattern matching and intuition - such as AlphaGo - have the potential to replace higher-skilled workers such as radiologists, tax advisers, and customer-service reps. Hassabis said DeepMind is "thinking very carefully about how to ethically use and responsibly deploy" the technology.
The concern that smart machines, built with the deep learning AI technologies showcased in AlphaGo, could displace masses of skilled human workers, indicates the magnitude of Google DeepMind's achievement. Once Alphabet starts deploying next-generation AI technologies in the marketplace, its stock is likely to soar.