Last year Google’s DeepMind defeated Lee Sedol a professional Korean Go champion 4-1. The result, hailed as a sign of AI, meant that terminator bots would visiting us soon.
But while impressive, the win wasn’t AI. DeepMind’s winning streak was the result of clever coding and a library of 30 million moves. Google overclocked DeepMind by adding CPU’s and taught it by leaving it to play itself. This let DeepMind mimic originality with learned moves and fast reflexes with computational grunt.
What DeepMind does highlight how easy it is now deliver automation using algorithms and cloud based computers. Arrive at any retail website and an annoyingly chirpy customer rep will offer you help. Using logic first seen in Eliza in the 1960’s, computers now answer simple questions like ‘can I have longer to pay’. Leaving real people to answer the tricky ones.
eProcurement benefits from these approaches. While the chatbot use-case is limited, deploying algorithms collect data and, compare and recommend supplier offers delivers clinical efficiency to buying decisions. Email based RFPs and spreadsheets are no match for repeatable scalable buying approaches.
If we were transported back to the 1970’s, to the locals our phones would look like Star Trek props. Using Google Maps (assuming the GPS worked) to would look like magic. For procurement that 40 year technology leap is unlikely to take as long to impact the business buying experience.