Exponential growth trends define current AI progress, according to Mustafa Suleyman. He argues that human intuition for linear progress fails to account for the rapid acceleration of machine learning. This perspective suggests that scaling laws remain robust. Practitioners should expect continued capability gains rather than a sudden plateau in model performance.