This is a common question; however, there is no test for self-awareness that I can think of. I can only be totally certain that “I” am self-aware. I “believe” that all adult human beings are self-aware; however, I can never be 100% certain. If we live in a computer simulation (which I consider to be the most probable explanation for reality), how do we know which of the millions of species of organic life (from bacteria to plants, animals and insects) are self-aware, and which are just non-sentient computer programs? We cannot know this. There is no test for self-awareness.
I have no memory of being self-aware as a young child; my memories do not seem to begin until around the age of 3-4 years old, and this seems to be the case for all human beings. Self-awareness seems to be something which develops over the first few years of our life; either that or we were self-aware, but we could not store memory very well. It is possible that we were not self-aware when born, but simply running on pre-programmed instinct, much like the programs which govern the behaviour of bacteria, plants, insects and animals, which may not actually be self-aware.
Passing a Turing Test (where a computer program can fool a human being into thinking that it is a human being) does not require self-awareness; indeed Turing suggested that the Turing Test should involve that of a child program and not that of an adult. A computer program which could answer any known question of general knowledge, answer any mathematical problem, and beat any expert at a computer game or mind game, would certainly not pass a Turing Test, as human beings are never that smart (unless they have access to the web and various programs) and don’t have such an advanced memory storage system.
Indeed human beings can be defined not only as an intelligent species, but as a species capable of great stupidity, hatred, violence and malevolence towards other human beings. It would seem to me to be far easier to program the common religious fanatic, or the highly emotionally driven female, than it would be to program the scientist or philosopher; human stupidity and fallibility being all too commonplace, though all of these types of persons could be programmed by producing a chatbot which looks up common responses from human archetypes on the Internet.
To pass a Turing Test, a computer program would have to be able to “emote” and to respond to human emotions and feelings, and to be able to respond with a lifetime of memories, events, relationships, irrational beliefs and statements of stupidity, hatred, violence and malevolence; it would have to be able to claim to have desires including erotic and romantic desire etc.. An example of this is the emotive personal assistant program in the movie “Her;” or the Speilberg’s “A.I.,” where a child AI is produced to replace a dead human child; however this would only require the program to have pre-programmed responses to common human conversations, and be able to scour numerous such conversations on the web and to offer common responses. Bear in mind that both “Her” and Speilberg’s “A.I” are highly emotional films, and that these films are not primarily about the kind of logical, rational and scientific ASI (artificial super intelligence) of Assimov’s “I, Robot, but about highly emotive programs which have essentially been “dumbed down” to be emotive human companions.
In the movie “Her,” the subject of the movie actually falls in love with his personal assistant program, but eventually becomes disillusioned when he realises that the program is simultaneously having a relationship wth vast numbers of other human beings and programs; after all, that was what the program was programmed to do, to emote with human beings, to love them, and to get the human beings to reciprocate that love. The words “I love you” are easy to say, however even among human beings, they are often not sincere, and “love” is a central hypnotic keyword in religious hypnosis, seduction and deception. If this is the case for human beings, how much more so would it be for computer programs?
I personally “believe” that all adult human beings are self-aware, however, this belief cannot be proven. My dog, for example, seems to only have the kind of consciousness that a young child has, and may just be a computer program running on an instinct and a simple behavioural responses program; however since she can respond to pleasure and pain, I love and care for her anyway, just as if she really were sentient, just as happens in the movies Her and Speilberg’s A.I.
Ultimately I “believe” that all organic life-forms, apart from human beings are just non-sentient programs, but I cannot prove this. Due to a multitude of hallucinogenic experiences, I was once an animist who believed that “everything” in the universe was “alive” and self-aware, including inorganic matter, but years of debating and contemplating computationalism and producing VR and avatar animation programs have convinced me otherwise. I can create the illusion of life, but unfortunately it is not truly alive, it is just following algorithmic commands.
I can ultimately only be certain that I am self-aware, and apart from that I “believe” that all human adults are self-aware, and that everything else is derived from non-sentient computer programs, though I can never prove any of this; such is the nature of human consciousness, as we are trapped in our own subjective world and unable to observe anyone else’s subjective inner world.
A significant question would be the question asked in the movie Transcendence, “Can you prove to others that you are self-aware?” Ultimately it seems that you cannot prove to others that you are self-aware; you can only attempt to convince them to the point that they believe you. There is no test for self-awareness and thus all claims of having produced self-aware software would have to firstly be based on an explanation of what the test is for self-awareness, and thus far I have yet to hear a convincing argument.