A recent research paper by Assistant Professor Xiang Ren and a Ph.D student Yuchen Lin shows that current AI technology lacks common sense.
"Current machine text-generation models can write an article that may be convincing to many humans, but they're basically mimicking what they have seen in the training phase," said Lin.
The researchers developed a tool that can essentially test the amount of common sense in a AI generate piece of text.
"We find even the strongest model, called the T5, after training with a large dataset, can still make silly mistakes." Lin added.
According to Ren and Lin, prior researchers had not tested the common sense ability of the artificial intelligence sufficiently.
For more on this story and research click HERE.