GPT-4o was presented as a model able to reason across audio, vision, and text. The rollout began with text and image capabilities; the new voice experience followed a separate availability path.
Why it mattered: quicker, more natural interaction suggested assistance could become less dependent on a typed chat box.
The perspective: an announcement can show the direction of a model while individual features roll out at different times.