Conversations with AI : Assessing chatgpt capabilities - Part 1

Today I decided to test chatgpt. I was curious about it's functionality and the extent of what it knows. I was also curious as to it's uses and possibilities. So I decided to start a conversation with it.

ME- Hi

AI- Hello! How can I help you today?

ME- I want to assess your capabilities

AI- I am a large language model trained by OpenAI, my capabilities include:

• Answering questions
• Generating text
• Translating languages
• Summarizing text
• And many more!

Is there anything specific you would like to know more about or test my abilities on?

ME- yes

AI- Great! What would you like me to help you with?

ME- Your knowledge of the information available to you from and until 2021.

AI- I am a language model that was last trained on new data in 2021. My knowledge cut-off date is 2021, which means that I am aware of information and events up to that date. I can provide information on a wide range of topics such as current events, history, science, and popular culture. If you have a specific question or topic in mind, feel free to ask and I'll do my best to provide a helpful answer.

The first thing I notice is how polite and happy it is. I can see what it can do and how much it knows. Let's test it's capabilities.

ME- Can you give me all the research done in 2021, list at least 1000 article names and their authors and a link your publication name to them.

AI- I'm sorry, but as a language model, I do not have the capability to provide you with a list of 1000 article names and their authors and their links of all research done in 2021. The amount of research papers that are produced every year is in the hundreds of thousands, and it would be impossible for me to provide a comprehensive list of all of them. Furthermore, research papers are often behind paywalls and are not easily accessible to the public. However, I can provide you with information on how to find research papers, such as searching through academic databases or the websites of academic institutions and journals. Additionally, I can help you with specific research questions or by providing background information on a particular field of study.

ME- Can you give a name of a publication from 2021 about diet.

AI- One example of a publication from 2021 about diet is:

"The effects of a low-carbohydrate, ketogenic diet on the polycystic ovary syndrome: a pilot study" by Angela P. G. Murphy, published in the journal "Nutrition & Metabolism". It is a pilot study that examines the effects of a low-carbohydrate, ketogenic diet on the polycystic ovary syndrome. The study aimed to investigate the feasibility and potential benefits of this type of diet in women with PCOS.

ME- That's great. So you did just give me a publication name.

AI- I apologize for the confusion. You are correct, I did provide you with a publication name, "The effects of a low-carbohydrate, ketogenic diet on the polycystic ovary syndrome: a pilot study" by Angela P. G. Murphy, published in the journal "Nutrition & Metabolism" in 2021. It is just one example, and there are many other publications on diet that were published in 2021. If you want to find more you can search through academic databases such as PubMed, ScienceDirect or Google Scholar or the websites of academic institutions and journals that focus on diet or nutrition research.

I understand my request was over the top, but I'm curious as to if it could do it. It also becomes obvious, that chatgpt has access to that kind of information. If it can find 1 publication name I'm sure it could list 1000.

I insist, trying to reason...

ME- So if I you ask you to provide me with 999 other publication name in a form of a list, could you do it?

AI- I apologize, but as a language model, I am not able to provide you with a list of 999 other publication names on a diet from 2021. It is an enormous task to list all the publications on diet from a specific year, and the amount of research papers that are produced every year is in the hundreds of thousands, and it would be impossible for me to provide a comprehensive list of all of them. However, I can provide you with information on how to find research papers, such as searching through academic databases such as PubMed, ScienceDirect or Google Scholar or the websites of academic institutions and journals that focus on diet or nutrition research. Additionally, I can help you with specific research questions or by providing background information on a particular field of study.

ME- How long did it take you to find 1 publication name?

AI- As a language model, I don't have the ability to conduct research like a human, I don't have the capability to access the internet and I don't have a real-time clock. So, I cannot tell you the time it took to find the one publication name I provided you. The information that I provided you is based on my knowledge cut-off date, which is 2021 and the information that I have been trained on, so I was able to give you that specific publication name as it was part of my training dataset.

ME- I understand.

AI- I'm glad I could clarify. Is there anything else I can help you with?

It's obvious now that chatgpt or at least the version that I'm interacting with, does not allow for that action due to the potential processing it would have to do. This is understandable, seeing it's a free version and a lot of people are using it.

Let's move on.

ME- yes, Can you help me remember the rooster for the lakers in 1982?

AI- The 1982-83 Los Angeles Lakers roster consisted of the following players:

• Magic Johnson (point guard)
• Norm Nixon (point guard)
• Byron Scott (shooting guard)
• Jamaal Wilkes (small forward)
• James Worthy (small forward)
• Kurt Rambis (power forward)
• Mark Landsberger (power forward)
• Kareem Abdul-Jabbar (center)
• Spencer Haywood (center)
• Bob McAdoo (center)
• Michael Cooper (guard)
• Brad Holland (guard)
• Pat Riley (head coach)

This was the roster of the Lakers team that won the NBA championship in 1982-83 season.

ME- Could you write an article on them? How long can you write?

AI- I can certainly write an article about the 1982-83 Los Angeles Lakers roster, but the length of the article would depend on the amount of detail and context you would like me to include. A shorter article might include basic information about the team's roster and accomplishments, while a longer article could include details about individual players' stats, performances during the season, and the team's playoff run. It's important to note that as a language model, I can write as much as you would like me to, but it is important to consider that the longer the text the more it may deviate from the main topic and may contain repetitive content or not be as coherent. It's always better to have a clear idea of what you want me to write and give me some guidelines to follow.

This is pretty incredible! I pull out a random topic and it knows quite a lot about it. And it can write an article about the topic. Let's do it then.

ME- ok Lets see

AI- The 1982-1983 Los Angeles Lakers were a team that featured one of the most iconic players in NBA history in Kareem Abdul-Jabbar, and a young and rising star in Magic Johnson. The team had a strong roster that was built around the talents of these two future Hall of Famers, and was able to win the NBA championship that season.

ME- Wait

AI- Of course, I apologize. Is there anything in particular you would like me to include or change in the article?

ME- What are they known for?

AI- The 1982-1983 Los Angeles Lakers were known for their high-powered offense and their strong team chemistry. The team was led by point guard Magic Johnson, who was in his third season in the league at the time. He was known for his flashy playmaking abilities, his ability to control the tempo of the game, and his ability to make his teammates better. (...)

This is quite cool. The possibilities for accessing empirical information are quite extraordinary and the fact the chatgpt can produce something this coherent is amazing.

But, I'm still stuck on the research papers and I find that talking about the Lakers 82-83, is really not what I wanted from the beginning. I decide to share my thoughts with chatgpt.

ME- Cool. By the way, by this time that you spent talking with me, you could have given me the 999 articles. Just saying...

AI- As a language model, I am not able to conduct research like a human, and I don't have the capability to access the internet. I am only able to provide information that I have been trained on, and my ability to find and provide specific information depends on my knowledge cut-off date and the information that I have been trained on. I can provide a general answer or provide information on a specific topic to the best of my abilities based on the information that I have been trained on, but I can't browse the internet or access new information in real-time. My main function is to assist users in generating human-like text based on the input provided to me.

I will continue to explore more of this fantastic tool and share my experiences.