Impact of voice recognition on digital environments and Digital strategy
- Mark Skilton
- Aug 22, 2014
- 3 min read
The growth of voice recognition is meeting a turning point in the market as accuracy rates are fast approaching good enough accuracies to enable both on-command and actual productivity enhancement better than traditional input devices. With accuracies being cited as up to 99% voice to text conversion (1) there is the realistic prospect of being able to interact with machine information.is the frontier of the Human – Machine interface.
I would argue though that this has to be seen in the context of human voice transcription to machine readable information and then how the machine, the computer actually understands what the instructions are and can meaningfully reply back in context. This latter step is hugely different to translation, its semantics and pragmatic understanding of the meaning to make sense and sensible action from this. We see evidence of this today in the voice activated queries on search engines and the ability to provide answers in context of the question being asked. Mobile device voice assistants like Apple Siri, Google Now and Microsoft Cortana suggest 83% to 86% voice recognition accuracy but often these are within a specific set of questions and parameters relating to the services the device can offer (2). This is different to the recent success of a computer program beating the Turin Test which involves a human asking random questions and not being able to tell it’s a computer answering. In June 2014,
“Eugene Goostman” the Chatbotcomputer program used keyboard input to successfully emulate a virtual 13 year old boy passing the origin October 1950 test set out in Turin’s paper “Computing Machinery and Intelligence”, Mind Journal (3) requiring a machine’s ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human (4)(5). It’s also questionable if Eugene passed this test in the sense of artificial intelligence and was more mimicry or what Turin might have subscribed to clever patterns from morphogenesis (6). This in no way reduces the great achievement and use of digital translation services, it’s a first major step to a fully integrated experience of humans and machines interaction.
Voice recognition is part of the emerging technology landscape of Human-machine integration journey. We can extend interfaces further into wearable technologies that sense our movements, body temperature and other data that is non-verbal in nature. Embedded technology may also create new possibilities to improve body organ function and improvement such as pace makers, hearing aids, proximity sensor guidance devices and eyesight augmentation. The root challenge beyond this is how to direct Human thought and commands to obtain and seek information as well as interact with others and other machines. Research is already enabling some brain-machine interaction including eye image recognition, thought pattern recognition and even emotion though recognition and control in a Brain-to-Brain interface (7) but it’s still early days. The possibilities are limitless.
Voice recognition today is a key area of emerging lifestyle and immersive technology that enables humans to work better with their living environment. It is a key stepping stone to providing accurate information requests to machines that can interpret, assist and augment our world. The big shift is that machine algorithms in speech recognition is becoming good enough to minimize errors and importantly reduce the time Humans spend in learning to use this voice recognition interface. Once it becomes reliable, easy to use, convenient and frictionless then adoption will accelerate as it becomes a better way than the old world of keyboards, symbol reading and handwriting. We are already seeing the shift to more in-context use of voice recognition and search like Health assisted living and hands-free mobile search. We are fast approaching that turning point. The design of digitally enabled immersive environments will change for ever.
Identity
Voice recognition
Service care
Voice assistance
Productivity
Dictation
Translation services
Assisted living
Healthcare
Voice commands
In Context queries
Voice searches
Augmented reality
Avatars
Gaming and multi-media
Interactive games
Movementassist
Self-driving cars and transport control
References
http://www.nuance.co.uk/for-individuals/by-product/dragon-for-mac/dragon-dictate/index.htm
http://www.cnet.com/uk/news/google-now-tops-siri-at-accuracy-says-analyst/
http://mind.oxfordjournals.org/content/LIX/236/433
http://www.theguardian.com/technology/2014/jun/09/eugene-person-human-computer-robot-chat-turing-test
http://blogs.wsj.com/digits/2014/06/10/did-eugene-goostman-pass-the-turing-test/
http://anilkseth.wordpress.com/2014/06/09/the-importance-of-being-eugene-what-not-passing-the-turing-test-really-means/
http://www.washington.edu/news/2013/08/27/researcher-controls-colleagues-motions-in-1st-human-brain-to-brain-interface/
Opmerkingen