The Path to Real Artificial Intelligence
First introduced in 1956, Artificial Intelligence (AI) has been so overused in the marketing of modern software, it is hard to know what is real #AI or just a good algorithm. Below are some analytics and technologies that have helped deliver on the promise of human-like intelligence from a computer.
#Correlation analytics is creating a rule based on a pattern.
#Heuristics advances correlation rules with interpretations and algorithms that look for more complex and incomplete patterns.
#MachineLearning is a way for algorithms to get smarter based on what they observe. They use modeling and make data-driven predictions.
#ArtificialNeuralNetworks was a short-lived architecture that mirrored our human brains’ structure but lacked the free-flowing connectivity data. (GPU’s have given new life to #neuralnetworks.)
#DeepLearning is extensively training the computer by repetitive pattern modeling to get the answer right on its own.
#NarrowAI is when computers perform specific tasks often better than humans — like image classification.
Authors Margherita Pagani and Renaud Champion in their soon to be released book, “Artificial Intelligence for Sustainable Value Creation” evaluated over 800 AI solutions across 14 industries to simplify the confusing AI landscape.
We care most when AI potentially impacts our livelihoods or pushes our ethical boundaries but the potential for improving our lives is no longer science fiction. AI helped in the development of the COVID19 vaccine and equitable distribution plan.
Given the choice to have a machine vs a human, to do a simple medical procedure — I will take the machine every time. While I’m not 100% ready to give up driving yet, largely because I enjoy it, I would not hesitate to have an autonomous driving option to help optimize my time during my post-COVID daily commute. AI will continue to improve and so will our understanding of its impact on our lives.
