Speech recognition (SR) first made its entrance on the commercial stage in the early 1990s, but until recently it has struggled to make an impression as a credible speech transcription solution. In contrast, digital dictation (DD) is a comparative latecomer to the market and has gained credibility more from its flexibility than by providing a high-tech transcription offering. However, this seems preferable to many law firms, who choose it over the pure SR option. SR solutions have continued to improve, yet consistent achievement of accuracy figures close to 100% still seems some way over the horizon.
Crossover between the two solutions is in the critical area of dictation and the way in which each one uses human interaction. DD’s transcription solution is fairly low-tech and makes no bones about its requirement for third-party human intervention to create text but, crucially, it can accept the dictation of the original speaker in a single pass without further input once dictation is complete. By contrast, SR technology requires the ongoing involvement of the original speaker to check for inaccuracies and amend them.
