That's probably relevant for your use case, but it's not really obvious that it is necessary for Google's demoed use case. As the article mentioned 90%+ of pizza shops were willing to take orders from their system, which is basically the main metric that matters for the calling on behalf of paying customers use case.
No argument that what you & Google have done is more challenging.
I am curious whether the media backlash against Google Duplex "tricking people" will impact your business. Whether this is a thing that real people will care about, or whether this is the media blowing things out of proportion.
In any case, I do find it more of a "Fun Fact" than anything really meaningful :)
I think it won't impact us. Currenty we are making ~10,000 phone calls per month and we have zero complaints regarding "tricking people". I won't call it tricking at all. It's just a new form of human-computer interaction we all dreamed of. We propose some set of rules guided by Azimov's three laws of robotics. For instance, an AI shouldn't warn you upfront it's an AI, but if you ask "are you an AI", it must tell you so, not trying to trick you, etc. This is how we train our Dasha.AI. We already have a lot of experience in the field. Since launch 6 months ago, Dasha.AI talked at least to 50,000 people. We have enormous dataset of human reactions. Anyone interested in helping us in accelerating world's transition to robot workers and, eventually, to the general AI, are welcome!
Comments
That's probably relevant for your use case, but it's not really obvious that it is necessary for Google's demoed use case. As the article mentioned 90%+ of pizza shops were willing to take orders from their system, which is basically the main metric that matters for the calling on behalf of paying customers use case.
No argument that what you & Google have done is more challenging.
I am curious whether the media backlash against Google Duplex "tricking people" will impact your business. Whether this is a thing that real people will care about, or whether this is the media blowing things out of proportion.
In any case, I do find it more of a "Fun Fact" than anything really meaningful :)
I think it won't impact us. Currenty we are making ~10,000 phone calls per month and we have zero complaints regarding "tricking people". I won't call it tricking at all. It's just a new form of human-computer interaction we all dreamed of. We propose some set of rules guided by Azimov's three laws of robotics. For instance, an AI shouldn't warn you upfront it's an AI, but if you ask "are you an AI", it must tell you so, not trying to trick you, etc. This is how we train our Dasha.AI. We already have a lot of experience in the field. Since launch 6 months ago, Dasha.AI talked at least to 50,000 people. We have enormous dataset of human reactions. Anyone interested in helping us in accelerating world's transition to robot workers and, eventually, to the general AI, are welcome!