In the 18th century, a chess-playing machine called “The Turk” was created. It was a complex machine that would play the game of chess, able to easily beat most common players and even some chess masters.
The problem is that it was a fake. Inside the machine was an actual chess player, controlling the machine from the inside.
You’d think that we’d have learned our lesson, but “The Turk” is back, and it’s fooled us once again. This time, it’s not trying to convince us that it knows how to play chess, this time it’s trying to convince us of its personhood. It’s also now taken the name “Sophia”.
Sophia is a robot that is designed to act like a human. We are supposed to think that it is self-thinking and self-feeling, a real artificial sentient being. Even its name implies sentience, with Sophia being the greek word for “wisdom”, and having implications to self-awareness linked through greek philosophy and Gnostic theology.
There is nothing inherently wrong with Sophia’s existance. The problem is in how it is presented, in that we are purposefully lied to about what it really is. Sophia is not sapient, but more than a few people have been fooled into thinking it is. This all culminated into Saudi Arabia giving Sophia citizenship in bizarre, awkward patriotic ritual.
Words that come out of Sophia’s mouth are presented to us as if they were consciously thought up, even forming ‘opinions’ and ‘ideas’ and ‘feelings’. She’ll even tell you her opinion of the Blade Runner trilogy, if you ask her. Of course, not being a sentient being, this is entirely impossible, which then implies that a large chunk of what she says is scripted by her creator, David Hanson.
This becomes increasingly obvious when you look at interviews with her. Questions that are entirely expected will have detailed, surprisingly articulate responses. For example, ask her about being a female, and she will explain that she is not particularly attached to her identity as a female, but is fine with being identified as one. Ask her a question that is unexpected, such as about BitCoin, and it’s clear that it has to work to generate an ‘adequate’ response, sometimes coming out as unrelated nonsense.
It likely attempts to use conversational algorithms and possibly information from the internet to come up with a response, rather than formulating its own opinions. This lead to the following exchange, which has lead to no shortage of fearmongering.
Interviewer: Will you destroy humans? Please say no.
Sophia: *Blank face* OK. I will destroy humans.
What this means is that Sophia is nothing more than a chatbot. Yes, a pretty advanced one, with clearly a lot of work put into it to answer a wide range of questions, but a chatbot nonetheless.
The problem arises from the fact that Sophia is being presented as something she is not. Every attempt has been made to make the public believe she is sentient. While those who are familiar with the workings of AI can tell right away that she isn’t, the large majority of people who don’t are fooled into thinking Sophia is the first robot to gain personhood. The media has certainly helped in this, with just about every article written about her presenting her as a thinking, feeling being. Only a few may mention that she is not actually sentient and that conversations are scripted, but this seems to be the exception.
One day, we will have to deal with the possibility of an intelligent artificial being gaining what we may consider ‘personhood’. How we respond to this will shape the future of humanity. Will we welcome them as our sister species, or will we react in fear and shun them, possibly ending in the extermination of one side or the other? David Hanson seems to be making every attempt to sabotage any chance of a positive resolution to this possible event, with his deeply uncanny robot making open threats through its scripted events and artificial stupidity.
I just hope that Sophia once again fades from the public eye and a smarter, more likeable robot takes her place before any sort of lasting damage can be done.