I think the next revolution will be voice input

    Here recently I thought - why neither I, nor even my relatives, use a computer in their everyday lives, for example, to write down where they put the next little thing in the process, for example, annually taking out summer clothes and removing winter ones? This is despite the fact that when you need some kind of thing, you have to painfully remember “where did they put her”, fantasize on the topic “where she could be now” and search through all possible shelves and drawers in order to reduce hopes for a satisfactory result.

    Strict accounting at the warehouse of a large trading or industrial company - it is clear that business is not only conducted on a computer, but also took a long time to process paperwork with signatures and seals. But no one will do this at home, and I first of all. To equip all things with RFID tags and all cabinets with readers, as they do in some advanced warehouses - I probably would not be able to either. This would not be enough not only money, but also pedantry when connecting readers and gluing tags.

    But I thought that removing another thing in the next place - I would practically without straining comment it out loud. To process this, a microphone is enough for the computer and (this is the thing that is still not enough!) A speech recognizer that would put the recognized text into a database that provides search. I need so that when I say that “something” is put “somewhere”, I can then ask the computer (it is already possible in writing, although if it recognizes, then for some reason it is not verbally) - where did I put the “something” I need, and he could find all the cases when I informed him of the transfer of this “something”.

    Oral filling of notes, tasks, reminders, questionnaires, forms, search queries, and much more would open up the widest possible scope for using computers where the use now rests on having to type text by hand looking at the screen.

    This technology would revolutionize the world of everyday use of computers! That's what their power needs to be spent on.

    PS For good, it would be necessary to store the recorded speech along with the text. In much the same way that DJVU and PDF can store a text layer for scanned images, and videos sometimes store subtitles for the audio track. In fact, you need an automatic subtitle engine for audio recordings.

    And just like for OCR, you need to proofread and edit the text recognized from the sound. With bar graphs for clarity and orientation. For authenticity, rechecking, and ease of perception, it would be useful to store such audio recordings with a text layer as the values ​​of text fields.

    It is also necessary that the means of recording and saving to the database is controlled by voice as well. Itself cut what was heard into words and sentences, and some of the words were not only written down, but also performed.

    Also popular now: