Raw LLM Responses

I saw a video about ChatGPT and the Turing test. And one of the questions was if ChatGPT has seen any movies recently. ChatGPT responded in a humanly manner but ChatGPT didn't specify the movie and when asked about what movie it has claimed to see; it responded with a classic "I am just a language model I do not have any experiences or thoughts" or what ever. And there is a massive problem I see here with deciding if an AI is conscious or not by asking it questions in a conversation; It most likely doesn't have any memories of experiences that could relate to us. But you could program it to have data of certain memories. But that is still a problem. But there is also a solution. What if an AI robot was like a baby? By this I mean that it has no memories but it has the capacity to be taught. So given enough time, maybe it has gone to a movie. And if it learned language it could most likely tell you all about the movie it saw. How are we gonna expect an AI to answer absolutely every one of our questions? We do not have compact while sizeable in data storage computer's that could hold all current data for an AI to have what we call memories. Even if you were to give an AI memories for every single possible normal human thing. And even if you could, how would it unique to that AI and not be from someone else? That's why creating an AI as a baby with the capacity to learn and be taught as it ages would be our best bet to achieve consciousness or simulate consciousness. This method... AI could obtain a unique individual identity on level with a human.