In the early 1990s, a number of companies started to offer software that was designed to convert spoken words into text that could then be displayed or printed. Among the early commercial offerings was a program entitled “Dragon Naturally Speaking”. In 1994, I tried an early version of Dragon in my workplace. After installing a microphone to my computer, I spoke several sentences rather slowly, and watched words appear on the computer monitor. Unfortunately, there were quite a few errors in the written text, and I chose to stick with keyboard entry.
Now, more than two decades later, Dragon NaturallySpeaking has been improved such that for many users with hand tremors (from E.T., Parkinson’s Disease, Dystonia, and other causes) it performs well enough that it can serve as a person’s primary method for entering text into a computer.
Most modern cell phones offer users the option to speak into their phone to generate an email, a text message (or iMessage), a memo, or a variety of other “text” inputs. Those with iPhones (the iPhone 4s or newer) or iPads can ask “Siri” a wide variety of questions by voice. Many newer automobiles are capable of responding to driver voice commands (tune radio to FM 107.9; turn on air conditioning, call home, etc.). The number of “things” that can respond to human voice commands continues to grow. All of these capabilities rely on speech recognition technology.
— Fred Berko