Centaurs 🔱

Combining human and machine intelligence.

2016 – 2018

 

 
 
 

New York City, May 1997


Sixth and final game in the rematch between world chess champion Garry Kasparov and IBM supercomputer Deep Blue.

 
 

Deep Blue famously wins the game, the first time where a computer outplays a chess grandmaster.

 
 
 

 
 

Leon, June 1998

The world's first championship of advanced chess organized by Kasparov.

In advanced (alt: freestyle or centaur) chess, humans of any expert level play assisted by technology such as hardware, software and databases.

Over the following year, a community of advanced chess players develops, designing various approaches of combining intelligence for chess playing mastery.

 
as disturbing as it was exciting
— Garry Kasparov describing the experience
 

 

Having evolved their methods, centaur players regularly start out-playing both chess grandmasters and supercomputers in 2005.

Teams today win by combining average skill players, mid-range technology and strong interfaces. By optimizing for an efficient connection between human and machine, teams can better leverage the characteristics of either.

 

When the smartest guy in the world speaks, you probably want to listen.

 

 

 
 
 

Centaur

AKA: Intelligence Augmentation (IA)

Human

Creative
Adaptive
Intuitive

Individual strengths of a human chess players:

  • being able to construct meaningful strategic plans, taking into account the long-term quality of moves, which even the fastest PCs cannot foresee;
  • being able to discriminate meaningful moves from the meaningless, without wasting time on deeply calculating the combinations which can be deemed meaningless at first sight;
  • being able to critically judge and analyze a chess game, plan, opening or endgame.

Machine

Consistent
Objective
Precise

Individual strengths of a computer chess program:

  • being able to calculate at a fascinating speed - on an average PC of today, a chess program is able to calculate a few millions of positions per second, making it tactically superior to any human in complex tactical positions;
  • having access to a database of millions of tried and thoroughly tested opening moves and variations, with the ability to retrieve information from such a database very quickly, and to store such a database on hardware resources available to most modern PCs;

Centaur Characteristics

(of advanced chess in particular, but centaurs in general)

  • increasing the level of play to heights never before seen in chess;

  • producing blunder-free games with the qualities and the beauty of both perfect tactical play and highly meaningful strategic plans;

  • giving the viewing audience a remarkable insight into the thought processes of strong human chess players and strong computers, and the combination thereof.
 

 

Examples

Human



Script writers, story tellers.

Centaur


Netflix
Content generation team uses big data to determine preferences of key demographics. Series have their themes, actors and story arcs suggested by algorithm. (More)

Machine



Big data on user behavior and quantified content metrics.



 

 

Defining Quotes

There are two kinds of jobs now - jobs that you can do better with an AI assistant, and jobs that an AI assistant can do better than you.

Scott Santens

IA, or Intelligence Augmentation, is all about empowering humans with tools that make them more capable and more intelligent, while traditional AI has been about removing humans fully from the loop.

Brad Neuberg


 

MIT Press (Feb 2018)
How To Become A Centaur

So, how do you find the best “+” for humans and AI?  How do you combine humans’ and AI’s individual strengths, to overcome their individual weaknesses?  Well, to do that, we first need to know exactly what humans’ and AI’s strengths and weaknesses are.

 

New York Times (May 2017)
Google’s A.I. Program Rattles Chinese Go Master as It Wins Match

Computer scientists say that often the best use of artificial intelligence is not to pit it against humans, but to pair it with them. To that end, two Go professionals, each teamed with AlphaGo, are scheduled to play against each other on Friday. Mr. Hassabis has said that top amateur Go players, with the help of AlphaGo, can generally manage to beat the software program in a match. In short, a human with a computer is still stronger than a computer.

 

Tim O'Reilly (July 2016)
Don’t Replace People. Augment Them.

"Those weavers who smashed machine looms in Ned Ludd’s rebellion of 1811 didn’t realize that descendants of those machines would make unbelievable things possible. We’d tunnel through mountains and under the sea, we’d fly through the air, crossing continents in hours, we’d build cities in the desert with buildings a half mile high, we’d more than double average human lifespan, we’d put spacecraft in orbit around Jupiter, we’d smash the atom itself! What is impossible today, but will become possible with the technology we are now afraid of?

 

Sydney J. Freedberg Jr. (November 2015)
Centaur Army: Bob Work, Robotics, & The Third Offset Strategy

“Human-machine collaboration is allowing a machine to help humans make better decisions faster,” Work said at the Reagan Forum. “That is a big, big difference. There is an artificial intelligence (AI) bias right now generally in the community [i.e. using unmanned systems to replace humans, rather than augment them]. But…automated systems use algorithms based on old data, [and] we’re up against a thinking adversary that is changing strategies all the time.”

 

Alvin DMello (October 2015)
Rise of the humans: intelligence amplification will make us as smart as the machines

"In contrast to AI, which is a standalone system capable of processing information as well as or better than a human, IA is actually designed to complement and amplify human intelligence. IA has one big edge over AI: it builds on human intelligence that has evolved over millions of years, while AI attempts to build intelligence from scratch.

 

The Economist (May 2015)
The dawn of artificial intelligence

"The best players in the world are not machines however, but what Garry Kasparov, a grandmaster, calls “centaurs”: amalgamated teams of humans and algorithms. Such collectives will become the norm in all sorts of pursuits: supported by AI, doctors will have a vastly augmented ability to spot cancers in medical images; speech-recognition algorithms running on smartphones will bring the internet to many millions of illiterate people in developing countries; digital assistants will suggest promising hypotheses for academic research; image-classification algorithms will allow wearable computers to layer useful information onto people’s views of the real world.

 

Mike Cassidy (December 2014)
Centaur Chess Shows Power of Teaming Human and Machine

Yes centaur — and Kasparov was apparently the first. Rather than half-horse, half-human, a centaur chess player is one who plays the game by marrying human intuition, creativity and empathy with a computer’s brute-force ability to remember and calculate a staggering number of chess moves, countermoves and outcomes. The centaur story is an elegant example of the way visionaries see the optimal interplay between humans and machines. Teaming the two in chess, experts say, produces a force that plays better chess than either humans or computers can manage on their own.

 

George Dvorsky (May 2013)
Humans With Amplified Intelligence Could Be More Powerful Than AI

The third step involves the genuine augmentation of pre-frontal cortex. This is the Holy Grail of IA research — enhancing the way we combine perceptual data to form concepts. The end result would be cognitive super-McGyvers, people who perform apparently impossible intellectual feats. For instance, mind controlling other people, beating the stock market, or designing inventions that change the world almost overnight.

 

Garry Kasparov (February 2010)
The Chess Master and the Computer

In what Rasskin-Gutman explains as Moravec’s Paradox, in chess, as in so many things, what computers are good at is where humans are weak, and vice versa. This gave me an idea for an experiment. What if instead of human versus machine we played as partners? My brainchild saw the light of day in a match in 1998 in León, Spain, and we called it “Advanced Chess.” Each player had a PC at hand running the chess software of his choice during the game. The idea was to create the highest level of chess ever played, a synthesis of the best of man and machine.

 

Further Reading

  • https://www.oreilly.com/ideas/building-human-assisted-ai-applications
  • https://magicalnihilism.com/2016/03/31/centaurs-not-butlers/
  • https://medium.com/basic-income/deep-learning-is-going-to-teach-us-all-the-lesson-of-our-lives-jobs-are-for-machines-7c6442e37a49#.bi7y33b5m

Sources

  • advancedchess.netfirms.com (archive.org)
  • https://en.chessbase.com/post/hydra-mies-the-quarter-finals-of-freestyle-tournament
  • https://en.chessbase.com/post/dark-horse-zacks-wins-freestyle-che-tournament
  • https://en.wikipedia.org/wiki/Advanced_Chess
  • https://en.wikipedia.org/wiki/Hydra_(chess)
  • https://en.wikipedia.org/wiki/Fritz_(chess)
 

Evidence against Centaur Chess superiority

Some argue that freestyle chess players had a window of success against machine players, and that AI has generally caught up with human chess players. 

 
 
9616585814_394869dc1a_o.jpg

Flickr