Using off-the-shelf AI without proper training often leads to unwanted hallucinations. This is why techolution is dedicated to establishing robust Human-AI partnerships using our RLEF AI platform: bit.ly/4ePO6n6 #AIDoneRight
Edge cases are hard… Notice how this Waymo self driving car can’t quite figure out why there’s a tree in the road in front of it… that keeps moving. Humans use inductive reasoning to rapidly process this nuance and make correct decisions. When AI doesn’t have enough “training material” to process this nuance it defers to rules based AI and uses deductive reasoning. AI sees the tree, knows that trees don’t grow on roads and concludes that it must circumnavigate the obstacle. However, the tree is moving, so AI proceeds through the “do-loop” again. AI won’t produce expected outcomes until you invest the time to train it just like you would a new employee. You wouldn’t take your college intern and turn them loose on your data, but organization after organization is doing this very thing with off the shelf AI products. No surprise, most are disappointed with the results. Want to learn more on how to get real world AI results? Check out techolution’s AI accelerators at https://lnkd.in/eKjHerfi #repost #AI #missionoutcomes
