Princeton profs plum secrets of Sudoku

Watching people puzzle over numbers, researchers glean info about associative memory.

Could Sudoku, that brain-teaser puzzle phenomenon that appears in newspapers everywhere, be the key to artificial intelligence? ABC News reports a new study suggests that playing Suduko exercises neural pathways. Scientists may learn about about how to make computers more intelligent as they study how people play Suduko.

Princeton University's professor John Hopfield, recently published a paper on the arXiv physics website which explores how the human brain solves a partial clue in numerical puzzle pattern by using associative memory.

Computers have the advantage of storing and processing large amounts of information but aren't capable of using associative memory. Hopfield theorizes that associative memory could be implemented in silicon chips.

"In neural terms, the signals developed ... can produce a strong and reasonably accurate feeling of correctness of the item retrieved," Hopfield says.

"This fact may account for our strong psychological feeling of 'right' or 'wrong' when we retrieve a memory from a minimal clue."

Associate professor Andrew Paplinski is an Australian computer scientist who specialises in neural networks at Monash University in Melbourne, says that being able to mimic associative memory would give computers "extreme robustness of pattern recognition."