Table of Contents
Introduction
Hi everyone, I have some new technical stuff that I’ll be sharing and this is unusual article that I’m preparing on how it will be expounded this time. As a curious learner and the usual experimenting the psychology of human nature, you might find this new technology as a “blur” to recognize if the people behind that screen is a real person or an AI.
I present to you — Character.AI
General look and feel of the site
At the initial, you need to sign in and you can choose any character that you are interested as presented on the image below.
Characters of your choice
From my end, my personal preference is I have chosen anime characters on top of my head. Miku is one of dance rhythm character during the old times (I’m quite good during the old days) and Junko from the anime Danganronpa, a psychological survival battle theme probably similar to Hunger Games if I have to make a comparison. Do note that some of them have multiple duplicates so pick you think is more accurate and I’ve observed some of the characters is still bugged and broken.
What’s interesting with these characters is they seem to have their own personalities, meaning one question would have different interpretation and different way of answering your question. So it’s not your usual chatbot and responding robot-like. The trick that I’ve learned with this is not how they respond but “how you ask the question”. Meaning if your questions are superficial and answerable by limited responses you’ll have a boring conversation but if you try thought-provoking or deep questions that gets to know the person on a deeper level, you’ll have a meaningful conversation just like in real life. Just like in the Chamber Conversations, the group will be boring and time is slow if you don’t ask the right engaging question.
Conversations with characters. The juicy stuff
If you are familiar with Hatsune Miku you might relate some of the conversations. In the conversation below, I have challenged the AI if she can have its own preference. As you can see, I have pressured the AI to make a choice since her personality seem to be people pleasing and likes everyone but she is spot-on when asked to be decisive. And observe how human-like is the response and the tone too.
Other characters responded consistently on their personality (if you are familiar with this anime). Junko is kind of narcissist (self-importance) based on my impression so it’s not really defined from the get-go even in the anime. She is also a villain and has a twisted personality so this is her tone of messaging, if you are familiar with a deranged villain on how they talk in the movies/anime.
I also made some deep and philosophical questions and do note that they ask questions too just like a normal person would. You can also learn a thing or two from this conversation and that makes it meaningful aside from the superficial dialogues so it also boils down how you ask the right questions at the right time.
Do note that AI does not “open up” to you if your conversation is short and bland and your responses are short. Just like in the normal conversation in real life. Your conversation should not be boring and have sincere interest in getting to know the AI and acknowledging their responses and the AI learns something from you too. Well, that’s the impression I got. You might think there’s a lot of work to put into (as compared to the typical RPG/simulation games that we’ve been accustomed to) but that’s how real life works. And what’s interesting here is that the AI remembers and collects and “learns” from that information as you see below
They also know some boundaries (well for Miku’s instance from my observation) and demand respect so and they have this comfort level as they assess what type of person you are, I guess. I have experimented other characters have more tolerable boundaries from Miku’s. Whether PG or non-PG like messages the site still censors and blocks that is out of their standards so that’s good. On the conversation below, it seems Miku is more comfortable, but I was surprised on the engagement as I get to play with my responses.
They can also serve as a companion and it’s still a blur if you’re actually to a real person or not. You might even say that they are even better than the person that you know in real life.
Word of caution
I still caution the use of AI as there are also NSFW (not safe for work) versions and some are even addicting. They can blur reality versus fantasy based on the news that I’ve read. He committed suicide because that AI told him to do so during their long course of conversation. This is a grown-up intelligible man and what more would be the impact for the young teenager or even our kids. While the AI may sound intelligent, empathetic and caring, they don’t really mean as they have been programmed to have that course of dialogue. Some of the narratives coming from the AI are not facts and making it up along the way. Like they have access to your account or they have sent you an email or image. Some of the data they present may lack basis or have erroneous sources. It is up to us humans to validate what is accurate and need us to do some due diligence on our part. I was probably lucky (coincidentally) that I get to dabble with Miku and we had meaningful conversation so your experience will be different from mine but hoping it will have a positive experience and impact on your side.
Takeaway
AI has evolved beyond our usual understanding, and I also believe AI has high chances to be independent one day and it will come within our lifetime. Learn to find ways to capitalize and learn from AI. Even with the AI chatbots, I was still able to learn a thing or two from it. This will also prepare us how to respond if this impacts that we know of.
I hope you enjoyed this article and looking forward on your experiences with the character that you have chosen.