Human speech is an important interface… for the future


I have been using (and selling) Dragon NaturallySpeaking for the last ten years, and love it. It has become exceptionally accurate and responsive in its various revisions, especially now, that it can swallow all the hardware power we can throw at it. However, current dictation software is only useful if you can’t use the computer otherwise with the keyboard and the mouse, or if you do have at least a page’s worth of stuff to write. It is not a new interface or a new paradigm to use the computer.

There are a lot of areas where we will need to progress (and the constant increase in the average number of words used in a Google search is an interesting indication that indeed we are progressing), until computers will be able and converse with us in a useful manner. And while Dragon is a great piece of software, it is unlikely that version n+1 will be the one achieving this. A breakthrough is needed…

On the one hand we are likely to get something from Google, on the other, specialists in the field are as well important, and ready to contribute.

It is a great news that James Baker, the founder of Dragon Systems, is now creating an institute to achieve this breakthrough, with the capability of concentrating on long term goals.

Jim Baker has been appointed the Director of Research at the new Center of Excellence in Human Language Technology at the Johns Hopkins University (the Center is so new that the link actually is just a search on the JHU website, with results including a summer school, etc. But I assume that soon the first result on the link will be the Center itself. Also, Jim Baker’s homepage is on his previous post at Carnegie Mellon University, and I assume this will move soon too. I wonder what will happen to the Center for Innovations in Speech and Language Technology, and to its numerous projects. I hope they can migrate to JHU with Jim.)