You may remember the recent post about the next Window 10 OS and about Siri and Cortana, the voice recognition software provided by Apple for their iPhones and Microsoft for their upcoming operating system for Windows phones and tablets.
Now I wonder how you feel about the idea of actually interacting with a device, or a machine for that matter, even if we have all seen it in the movies for years? I don’t mean the Dick Tracy type wrist communicator, but speaking to your mobile phone and getting an answer which shows you that your question or request was correctly understood and having the knowledge to reply correctly.
You may already be familiar with voice recognition software if e.g. you are using ‘Dragon Naturally Speaking’ for writing on your computer, or perhaps doing a language course where for correcting pronunciation you speak to the microphone and get the corrected sound if need be.
The latter type does not require the software to do other than recognize the words used by you, there is no requirement for it to think of a reply. However, it can be really frustrating to teach the software to interpret your spoken words correctly all of the time.
Siri, and quite likely Cortana too, will have had the benefit of extensive speech recognition technology breakthroughs so that we can now expect huge improvements in the functioning of these exciting tools and what has been achieved in the past couple of years is more than in the previous 15 years.
It has been possible with the enormity of usage data available to the companies developing voice recognition and the more the merrier which means better algorithms delivering better accuracy.
The thing is that voice interface with your computing device is more natural than clicking or touching and so we are entering the realm of AI or artificial intelligence and the major players like Apple, Microsoft and Google, Baidu [China] and Amazon, are all putting up a lot of money to make Search using voice recognition.
You can see where this takes us. Imagine the kind of data needed to understand not just various languages but dialects and inflections and nuances in word usages.
Yet we know that machines already can see better, recognise objects, listen and hear better. How long will it be before our computer or tablet listen to us 24/7, and as a result will know your needs before you do. Already we see major makers of appliances providing what is called a ‘connected home’, mostly voice activated.
One might speculate whether – or rather when – the AI systems will have enough human knowledge to enable it to answer all your questions, or at least be better at it. After all, human knowledge is not infinite and science says that between 100 to 500 billion concepts might approach the full extent of useful human knowledge. Bet you didn’t know that and we won’t tell the AI. The aim is to bring a generalized intelligence into focus.
This may sound rather worrying and there will be plenty of fear mongering among the luddites who might believe that the Sci-Fi versions of artificial intelligence will take over with unimaginable outcomes. But with a bit more realistic thinking one recognizes that we are a long way off such a reality, indeed AI becoming self-aware is not a threat for decades to come and any form of domination by robots with AI. The truth is that the kind of higher intelligence to rival human ditto is still a long way off.
OK, so I am not saying if cannot happen, chances are it will, but it is speculation and if such high-level intelligence emerge from machine systems over the next decades then hopefully we have enough sense to enter into serious debate over the best way to ensure we have the means to prevent any kind of apocalypse.
Meanwhile enjoy the advantages of being able to interact with your Siri or Cortana or voice operated gadgets, it makes life easier for you and presents no threat. We at SeniorNet may have a problem if everybody in the class chats with their computer and the poor Tutor cannot get a word in!