14
$\begingroup$

I talk about the robot from: Hanson Robotics, which was granted the right to citizenship from Saudi Arabia.

I have found the following articles:

Your new friend is a humanoid robot

Source: theaustralian.com.au

Like Amazon Echo, Google Assistant and Siri, Sophia can ask and answer questions about discrete pieces of information, such as what types of movies and songs she likes, the weather and whether robots should exterminate humans.

But her general knowledge is behind these players and she doesn’t do maths. Her answers are mostly scripted and, it seems, from my observation, her answer are derived from algorithmically crunching the language you use.

Sometimes answers are close to the topic of the question, but off beam. Sometimes she just changes the subject and asks you a question instead.

She has no artificial notion of self. She can’t say where she was yesterday, whether she remembers you from before, and doesn’t seem to amass data of past interactions with you that can form the basis of an ongoing association.

Questions such as: “What have you seen in Australia?”, “Where were you yesterday?”, “Who did you meet last week?” and “Do you like Australia?” are beyond her.


Why Sophia the robot is not what it seems

Source: smh.com.au

You can often fool this sort of software by introducing noise. That could be literal noise – machines aren't great at filtering out background noise, as anyone with a hearing aid will tell you – or it could be noise in the sense of irrelevant information or limited context. You could ask "what do you think of humans?" and then follow up with "can you tell more about it?" The second question requires the robot to define "it", remember what it said last time, and come up with something new.

In the case of the ABC interview, the questions were sent to Sophia's team ahead of time so they were possibly pre-scripted. Just like an interview with a human celebrity!


Pretending to give a robot citizenship helps no one

Source: theverge.com

Sophia is essentially a cleverly built puppet designed to exploit our cultural expectations of what a robot looks and sounds like. It can hold a stilted conversation, yes, but its one-liners seem to be prewritten responses to key words. (As Piers Morgan commented during an interview with Sophia, “Obviously these are programmed answers.”)


Here are a few more references.

$\endgroup$
0

1 Answer 1

13
$\begingroup$

Sophia uses ChatScript.

You can read about what ChatScript can do here.

ChatScript keeps track of conversations with each user; can record where it is in a conversational flow and what facts it has learned about a user (you have to tell it what facts to try to learn). You can optionally keep logs of the conversations (either on a ChatScript server or locally on a freestanding device).

$\endgroup$
1
  • 1
    $\begingroup$ Thanks Rebecca! The link you provide talks about the Loving AI project and I think I've just found another github repo which is quite relevant: github.com/opencog/loving-ai. Sophia is mentioned several times if one does a search in the source code. $\endgroup$
    – tgogos
    Commented Oct 31, 2017 at 14:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .