Text-to-speech turns written words into spoken audio. Modern systems produce voices that sound natural and expressive, with realistic rhythm and emotion, a long way from the flat, robotic speech of the past, and they can do so quickly and in many languages.
Natural synthetic speech powers assistants, audiobooks, navigation and accessibility tools that read content aloud, and it helps people who cannot speak. The same realism enables misuse through convincing fakes, which has raised interest in consent and in marking audio as synthetic, so the technology is both widely useful and a reason to be more careful about trusting a voice.