The short answer is a lot. But how can i say that. First of all knowing how popular it is and how much power google has, it is only logical that it will keep improving. Additionally, in a recent study, it was demonstrated that neural networks (that google has certainly used before) can be trained to help accelerate language reproduction from text to speech with even very limited resources.
In the last google event, the google assistant demonstrated that it can call as your personal assistant, sounding exactly like a human, with all the right and perfectly imperfect pauses and “uuummm”s. It placed a call to make an appointment and although it is a demo, it was very convincing and a lot of people were scared that if this technology exists how can you know that you are talking to a real person on the phone?
Imagine having the ability to use google’s servers as a virtual neural network to quickly make a voice assistant based on your voice. I mean, from the studies published and the way the industry progresses such a thing may be one or two years ahead.
At that point i am not sure what the best choice is. Because if you end up regulating such assistants heavily, you will end up with people not using them or not taking them seriously. If they have to disclose it for example, the attitude of the person on the other end of the phone will change a lot. But we will have to wait and see what happens with such technologies. Because eventually such methods will be used by individuals, “hackers” and enthusiasts and how this will end up is very uncertain.
What do you think. Feel free to leave your opinion in the comments below the post and have a discussion.