Speech to Text Support in OLPC

[James Cameron]

Opinion follows.

This will be interesting.

Children learn to talk in whatever way meets their needs.  Before
talking, they learn the exact pitch to annoy the heck out of their
parents ... not sure what the mechanism is, probably just trial and
error, over days and months, but they end up generating sound that
matches the hearing of their caring adults.

A friend of mine has degraded hearing and a cochlear implant, her
hearing frequency response is very narrow.  Her children have learned to
speak within that frequency band and accentuate the sounds that she can
hear, suppressing the sounds that she cannot.  Her children sound
strange to others when they do this.

With a speech to text capability on the XO, no matter how bad it is, the
result may be that children learn to speak in exactly the way that
yields the best result.

A whole generation will sound different when we meet them.  Fantastic.

This also means it might not be necessary to have the speech to text
capability work with existing version adults.

-- 
James Cameron    mailto:quozl at us.netrek.org     http://quozl.netrek.org/