In an earlier S45 post mentioned (https://goo.gl/ordtuy) how one #artificialintelligence has the intelligence of a four year old human.

This current post is about a different AI showing how it can complete a math SAT exam, according to the Washington Post (22 Sept 2015).

Washington Post wrote: "...the race to create an autonomous, intelligent computer is still ongoing. The latest potential step forward comes from the Allen Institute for Artificial Intelligence (AI2) in collaboration with the University of Washington, where an AI system has gotten a fairly average score (for a high schooler) on the math section of the SAT -- 500 points out of 800. That's about 49 percent accuracy, and comes quite close to the high school senior average of 513 points." https://www.washingtonpost.com/news/speaking-of-science/wp/2015/09/22/ai-can-now-muddle-its-way-through-the-math-sat-about-as-well-as-you-can/

The two AI systems in question are GeoS and ConceptNet (http://conceptnet5.media.mit.edu/).

If you didn't know, the RSS feed for S45 is here, to ensure you don't miss posts: https://feeds.feedburner.com/GooglePlusSingularity2045
Shared publiclyView activity