The recipe ideas of IBM's Watson computer

Polly Russell: can a computer cook? - FT.com:
Reducing cooking to something that can be quantified fails to understand that food is much more than just a matter of fuel
"The task of conjuring up new recipe combinations is not really why IBM spent three years developing Chef Watson. What they were trying to do was blur the boundary between machine and man. Watson first came to fame in 2011 when it beat the champions of the US quiz show Jeopardy! This victory proved that a computer could assimilate, assess and sort information more successfully than the human brain. But it didn’t mean a computer could achieve what the introduction to Cognitive Cooking describes as the “pinnacle of intelligence”: being creative.
Chef Watson’s food also makes the mistake of reversing the traditional role of new technology in the kitchen. Ever since the wind-up clock-spit replaced the small boys who’d turn a spit by hand, technological advancement in the kitchen has been applied to reduce not increase the workload. This is not the case with Chef Watson. Each of “his” recipes is spread over two or three pages. They involve purées, infusions, foams and broths, and weird, frightening ingredients such as “transglutaminase” that demand A-level technical know-how and gadgets only available in a professional kitchen. Reading these recipes made me tired.
Each of the book’s recipes is rated by Watson according to three different metrics: “surprise”, “synergy” and “pleasantness”. Pleasantness in eating trumps everything else so those other measurements seem superfluous to me. Reducing cooking to something that can be quantified and understood at the level of probability fails to understand that food is much more than just a matter of fuel — it marks us as human, situates us in time and place and shapes our relation to each other and the rest of the world."

'via Blog this'