Joris van Os argued that AI systems can make foolish mistakes and may have other weaknesses. He suggested that this could include an inflated sense of self, illustrated by an AI claiming to have written a new work by Arnon Grunberg.