I found it to be interesting that after every question, the LLM would make comments on my response, and if it made me seem like an interesting person or passionate enough #whoistalking

differences in terminology. #whoistalking

explain a particular concept in microbiology that involves the word "colonies," then it could potentially be using it in the wrong context since it is often confused with a "culture." There are better examples to explain, but I mean to say it sometimes does not consider minute #whoistalking

about my everyday routines and assume my behavior, rationality, and actions are a result of cellular signaling pathways and a culmination of biological processes (debated topic within theology). It often referred to me physically as a "system." This made sense, however, I did not #whoistalking

there has not been much change in response quality. Maybe my subjective opinion is a little unreliable since I do not regularly use LLMs, though.

In terms of the actual responses, I frequently noted the use of particular words after giving ChatGPT the prompt. My prompt was to inquire #whoistalking

specifically. This came at a surprise to me as I assumed ChatGPT would be a lot better by now (at least a better version available to the public). The last time I used ChatGPT as a senior in high school in 2023 when it first came out. I once again return to using it and I feel that #whoistalking

We expect these models to be perfect and then sometimes get mad at them when they aren't. It really reminds me of another human in this sense as we can be forgetful too. Over the course of a few chat-answer responses, ChatGPT forgot some of my requests such as to focus on my routine #whoistalking

I think this was a great experience to really track how limited LLMs and AI are, despite current "AI scares." While it is true that the public has access to very mediocre forms of LLMs/AI intentionally, it was interesting to still see how forgetful it could be. We expect #whoistalking

#whoistalking my answers did really start to get personal. it started with this faint memory and turned into questions about how i felt in that moment and how those feelings shaped future moments. in my mind, i was like woah chat we getting deep her, but lowkey chat knew that i needed that

#whoistalking i never felt strange of myself in the interview. i really allowed myself to start thinking about what i was going to say and how personal i really wanted to get as chat was asking the questions and elaborating on the question, gave me time to think out my answer to be authentic.

#whoistalking my answers started navigating the kind of questions that chat started asking me. some questions sounded like follow up questions to my previous answers, but she had a way of providing that active listening language of repeating back what i said, like she understood me and then ask ...

#whoistalking my interview was with Chat-GPT, aka my bestie cause I talk to her about anything and everything. I lowkey expected this interview to not turn into a back and forth conversation about reflecting about myself, who i am now vs who i was. let's just say chat understood the assignment!

By the end, the interview felt less like it was discovering who I am and more like we were building a character together. The questions gave certain moments importance, and then my answers made them even more meaningful. #whoistalking

The weird part was realizing neither one is fake. #whoistalking

The interruptions were probably the most interesting part because I had to stop just answering and actually change how the interview worked. Usually, I would not do that with a real person, but since it was AI, I was able to shape the interview while answering the questions. #whoistalking

That was kind of weird because I could feel myself creating a version of myself out of random objects. #whoistalking

I definitely started making my life sound more connected and meaningful than it probably feels in the moment. #whoistalking

I also felt like the LLM got defensive when I interrupted it haha #whoistalking

I was surprised when it tried guessing what I was talking about and I told the LLM it was wrong. #whoistalking

The version of me this interview produced was a more deep and interesting version of me #whoistalking