Columnists
Extracting my thoughts
I clipped an article out of the New York Times a few weeks ago that staggered me.
Scientists in San Francisco have developed a ‘prosthetic voice’ or ‘virtual vocal tract’ that has the potential to enable people who cannot speak to speak. All a patient will have to do is ‘think out’ what he or she wants to communicate, and the voice will do the rest. The research that prompted this breakthrough, also from San Francisco, found that there are differences in the brain regions that produce the sound of speech and the areas that control vocal movements. It focused on the latter. The study built on this finding by recreating the sound from the movement.
The prosthetic voice was tested in five epilepsy patients, who were chosen because they had already had electrodes implanted in their brains in preparation for neurosurgery to control seizures. All otherwise retained the ability to speak. The patients recited hundreds of sentences for the scientists, who recorded them and watched how neuronal patterns in the motor cortex were matched with movement in the tongue, lips and larynx.
The researchers then used the principles of linguistics to reverse engineer the vocal tract movements needed to produce those sounds, such as pressing the lips together here, tightening vocal cords there, shifting the tip of the tongue to the roof of the mouth, and then relaxing it. This detailed mapping of movement to sound was then fed to a decoder that transformed brain activity patterns produced during speech into movements of the virtual vocal tract, and a synthesizer that converted these vocal tract movements into a synthetic approximation of the participant’s voice.
The study subjects then just had to ‘think’ what they wanted to communicate. The result: slurred, but nonetheless recognizable speech at a natural pace (up to 150 words a minute).
(At present, some people with severe speech disabilities learn to spell out their thoughts letter-by-letter using assistive devices that track very small eye or facial muscle movements. However, producing text or synthesized speech with such devices is slow—typically about 10 words a minute.)
The implications: enormous. Presumably, the need to pre-capture vocal patterns from patients will evaporate as a bigger speech database is generated and the algorithms get smarter; meaning in turn that it can help people who have already lost their voices as well as people in danger of losing theirs. A wide group of conditions stand to benefit from this development: stroke, traumatic brain injury, and neurodegenerative illnesses such as such as Parkinson’s. MS and ALS are the examples cited by the researchers.
That is all well and good, as we are talking about the ability to assist a person to extract communication that would otherwise be trapped. But there are also technologies being developed that will extract my thoughts from me even though I would prefer to keep them private. This I don’t like so much.
For example, there are products available that can determine, with a proven accuracy level a lot higher than human judgment (60 to 75 per cent accuracy for the products versus 54 to 60 per cent for human judgment) whether I am lying, based upon the contortion of my facial muscles or the movement of my eyes. While they have their most immediate and obvious uses in the courtroom, they are starting to be applied in such fields as border security and job interviews. And just think if someone were to be able to coerce me into having electrodes strapped to my brain: my thoughts would be directly fed into a machine that could almost unerringly divine my truthfulness or untruthfulness in no time flat.
Once a product can tell whether I am lying or not, it is only a hop, skip and jump away from it determining just what I am thinking about. This may be awkward if I am being outwardly polite to someone, but keeping my inward sour thoughts about him or her to myself.
If I can’t take any consolation from the privacy of my thoughts, what substantive privacy do I have any more? I might as well just learn to accept thet my inner and outer thoughts have to merge. Which would be a shame, because I would never be able to interact with people I don’t like. And I might discover to my chagrin that there are many more people out there who don’t like me than I thought there were.
When all is said and done, let’s use this week’s column to toast scientific innovation, with the proviso that the risks always have to be monitored closely. Those are my extracted thoughts.
David Simmonds and Squib Ink will be taking a much deserved break. Don’t fret, David and his column will return in the September 4 edition of The Times.
Comments (0)