The original 1950s models had the vocabulary of 10 words. Versions thirty years later could comprehend more than Twenty thousand different words.
Voice recognition computers were in a position to recognize natural spoken language in the late 1980s and early 90s due to their increased storage and processing power. Recent examples of voice recognition technologies which you’ll find on your phone today do not depend on your processor or a restricted technology. Instead, they rely on the cloud’s database, which stores millions upon millions of phrases and words.
The addition of context is another breakthrough which allows it to provide more accurate outcomes. Rather than responding to a single word, the latest technology in voice recognition also incorporates surrounding words and phrases. By using complex mathematical models it uses probabilities to give the most pertinent results for a search.
However the new technology can bring additional problems for customers. Most important is the issue of customer privacy. While having a device that will learn about a person’s habits and patterns of speaking may help tailor the experience they have, it could cause doubts as to how their data is stored as well as different ways in which tech companies might be using their information. qb65r8uks1.