Analysis Insight (AI)

There are three things driving the rapid advances in artificial intelligence these days: Fast, inexpensive, and plentiful computers (i.e., speed). Large-scale AI computations had a one-million-fold increase in size from 1960 to 2010 (50 years). From 2011 to 2023 (12 years), another factor of one million has happened! Lots of data (i.e., storage/memory). The amount of data being generated per year is astounding. In 2024, it averaged out to be 19 trillion bytes per person per year! Data analysis (i.e., learning). Imagenet is a standard data set used to help computers identify objects in a picture. In 2010, computers misidentified objects 28% of the time. With new data learning technologies, by 2017, the error rate fell to 2%. Humans make 5% errors. The current advances in AI are mostly based on searching a massive number of scenarios, guided by astoundingly large data sets, and informed by sophisticated algorithms. Is this really intelligence?