Turning speech into text has become markedly more accurate and now runs quickly, often on the device. This powers live captions, voice control and dictation, and it is especially valuable for people who are deaf or hard of hearing, or who find typing difficult.
Real-time captions on calls and videos, and reliable voice commands, make everyday technology more usable for more people. Accuracy still drops with heavy accents, background noise, specialised vocabulary and less common languages, so work continues to widen coverage, but the gains have already turned a once-frustrating feature into a dependable everyday tool.