David Silver [the creator of AlphaZero] hasn’t answered my question about whether machines can set up their own goals. He talks about subgoals, but that’s not the same. That’s a certain gap in his definition of intelligence. We set up goals and look for ways to achieve them. A machine can only do the second part.

So far, we see very little evidence that machines can actually operate outside of these terms, which is clearly a sign of human intelligence. Let’s say you accumulated knowledge in one game. Can it transfer this knowledge to another game, which might be similar but not the same? Humans can. With computers, in most cases you have to start from scratch.

{ Gary Kasparov/Wired | Continue reading }

photo { Kelsey Bennett }

After 4 hours of training, AlphaZero became the strongest chess entity of the planet with an estimated ELO of around 3,400.

{ AlphaZero vs Stockfish 8 | ELO ratings of chess players }

more { How AlphaZero quickly learns each game [chess, shogi, and Go] to become the strongest player in history for each }

related { The ability to distort reality has taken an exponential leap forward with “deep fake” technology. We survey a broad array of responses. | Previously: Researchers can now detect AI-generated fake videos with a 95% success rate }

It would be difficult to be strong at chess if you had a subnormal IQ, but you certainly don’t need an IQ of above average. I’m sure you could find very strong grandmasters with IQs around about the 100 mark, which is the average. […]

What I have noticed in very strong players, though, is an extraordinary degree of concentration. You really do have to concentrate very hard for long periods. There is a very boring phrase for that, which is hard work. That’s often underestimated, while the idea of effortless genius is greatly overestimated.

{ Dominic Lawson/The Browser | Continue reading }

On May 11, 1997, chess grandmaster Garry Kasparov resigns after 19 moves in a game against Deep Blue, a chess-playing computer developed by scientists at IBM. This was the sixth and final game of their match, which Kasparov lost two games to one, with three draws.

Kasparov, a chess prodigy from Azerbaijan, was a skillful chess player from childhood. At 21, Kasparov played Anatoly Karpov for the world title, but the 49-game match ended indecisively. The next year, Kasparov beat Karpov to become the youngest world champion in history. With a FIDE (Federation International des Echecs) score of 2800, and a streak of 12 world chess titles in a row, Kasparov was considered the greatest chess player in history going into his match with Deep Blue. (…)

Kasparov first played Deep Blue in 1996. The grandmaster was known for his unpredictable play, and he was able to defeat the computer by switching strategies mid-game. In 1997, Kasparov abandoned his swashbuckling style, taking more of a wait-and-see approach; this played in the computer’s favor and is commonly pointed to as the reason for his defeat.

{ History | Continue reading | IBM archive of Kasparov vs Deep Blue }

By most all accounts a brilliant mind, Fischer was perhaps the most visionary chess player since José Raul Capablanca, a Cuban who held the world title for six years in the 1920s. Fischer’s innovative, daring play — at age 13, he defeated senior master (and former U.S. Open champion) Donald Byrne in what is sometimes called “The Game of the Century” — made him a hero figure to millions in the United States and throughout the world. In 1957, Fischer became the youngest winner of the U.S. chess championship — he was just 14 — before going on to beat Spassky for the world title in 1972.

But Fischer forfeited that title just three years later, refusing to defend his crown under rules proposed by the World Chess Federation, and he played virtually no competitive chess in ensuing decades, retreating, instead, into isolation and seeming paranoia. Because of a series of rankly anti-Semitic public utterances and his praise, on radio, for the Sept. 11, 2001 attack on the World Trade Center, at his death, Fischer was seen by much of the world as spoiled, arrogant and mean-spirited.

In recent years, however, researchers have come to understand that Bobby Fischer was psychologically troubled from early childhood. Careful examination of his life and family shows that he likely suffered with mental illness that may never have been properly diagnosed or treated.

{ Miller-McCune | Continue reading | Donald Byrne vs Robert James Fischer, “The Game of the Century,” 1956 | select Java Viewer and press set }

In the 1940s, the Dutch psychologist Adrian de Groot performed a landmark study of chess experts. Although de Groot was an avid chess amateur – he belonged to several clubs - he grew increasingly frustrated by his inability to compete with more talented players. De Groot wanted to understand his defeats, to identify the mental skills that he was missing. His initial hypothesis was that the chess expert were blessed with a photographic memory, allowing them to remember obscure moves and exploit the minor mistakes of their opponents.

De Groot’s first experiment seemed to confirm this theory: He placed twenty different pieces on a chess board, imitating the layout of a possible game. Then, de Groot asked a variety of chess players, from inexperienced amateurs to chess grandmasters, to quickly glance at the board and try to memorize the location of each piece. As the scientists expected, the amateurs drew mostly blanks. The grandmasters, however, easily reproduced the exact layout of the game. The equation seemed simple: memory equals talent.

But then de Groot performed a second experiment that changed everything. Instead of setting the pieces in patterns taken from an actual chess game, he randomly scattered the pawns and bishops and knights on the board. If the best chess players had enhanced memories, then the location shouldn’t matter: a pawn was still a pawn. To de Groot’s surprise, however, the grandmaster edge now disappeared. They could no longer remember where the pieces had been placed.

For de Groot, this failure was a revelation, since it suggested that talent wasn’t about memory – it was about perception. The grandmasters didn’t remember the board better than amateurs. Rather, they saw the board better, instantly translating the thirty-two chess pieces into a set of meaningful patterns. They didn’t focus on the white bishop or the black pawn, but instead grouped the board into larger strategies and structures, such as the French Defense or the Reti Opening.

This mental process is known as “chunking” and it’s a crucial element of human cognition. As de Groot demonstrated, chess grandmasters automatically chunk the board into a set of known patterns, which allow them to instantly sort through the messy details of the game. And chunking isn’t just for chess experts: While reading this sentence, your brain is effortlessly chunking the letters, grouping the symbols into lumps of meaning. As a result, you don’t have to sound out each syllable, or analyze the phonetics; your literate brain is able to skip that stage of perception. This is what expertise is: the ability to rely on learned patterns to compensate for the inherent limitations of information processing in the brain. As George Miller famously observed, we can only consciously make sense of about seven bits of information (plus or minus two) at any given moment. Chunking allows us to escape this cognitive trap.

{ Wired | Continue reading }

Competitors playing a match against Bobby Fischer, perhaps the greatest chess player of all time, often came down with a mysterious affliction known as “Fischer-fear.” Even fellow grandmasters were vulnerable to the effect, which could manifest itself as flu-like symptoms, migraines and spiking blood pressure. As Boris Spassky, Mr. Fischer’s greatest rival, once said: “When you play Bobby, it is not a question of whether you win or lose. It is a question of whether you survive.”

Recent research on what is known as the superstar effect demonstrates that such mental collapses aren’t limited to chess. While challenging competitions are supposed to bring out our best, these studies demonstrate that when people are forced to compete against a peer who seems far superior, they often don’t rise to the challenge. Instead, they give up. (…)

Because his competitors expect him to win, they end up losing; success becomes a self-fulfilling prophecy. (…)

In the early 1960s, the psychologist Sam Glucksberg demonstrated that the same effect could also inhibit creativity. He gave subjects a standard test of creativity known as the Duncker candle problem. The “high drive” group was told that the person solving the task in the shortest amount of time would receive $20. The “low drive” group, in contrast, was reassured that their speed didn’t matter. Here’s where the results get weird: The subjects with an incentive to think quickly took, on average, more than three minutes longer to find the answer. Experiments like this have led Ms. Beilock to conclude that people should be skeptical of evaluations based on a single high-stakes performance.

{ Wall Street Journal | Continue reading }

…while politicians occasionally use poker terms when discussing strategy, more of them–and more journalists—put on their game faces with terms from chess.

You would think that people would use that terminology the way chess players use it. Most of the time, however, they’re using the terms colloquially, even though they are using them in a strategic context. While that usage isn’t wrong, it’s not as precise as it could (or should) be. (…)

In chess, a “gambit” is an opening move, one that almost always sacrifices a piece, usually a “pawn.” But its more common use, one sanctioned by most dictionaries, refers to any risky or surprising strategic move: “Senate Majority Leader Harry Reid’s gambit to include a government-run insurance option in health care legislation has given a fresh tailwind to the idea despite opposition from conservatives,” one news report said.

{ Columbia Journalism Review | Continue reading }

photo { From Russia With Love, 1963 }