I’ve been meddling around ChatGPT for a week or so now, and the feeling is mixed. Not that it has not completely blown me away with its capabilities, but I’ve come to the conclusion that human intelligence might not be as valued from now on. I mean, this tool is giving me accurate answers to almost all the questions I have raised. It is writing essays and articles, paraphrasing documents, and even giving legal advice. I look at it with the same enthusiasm I got when I first discovered Google, but this is way beyond that. This is bigger than a human-like robot like Sophia or whatever Boston Dynamics is doing. This could be the source. This could become SkyNet!
Now that I have an AI assistant to write articles, I am thinking of doing this more often. This will be my first article in the series where I asked the AI to write things for me. Of all things, laziness is what will end us all.
My first question to ChatGPT was to write a starting paragraph of an article for my website for the topic ‘I asked an AI for an FAQ of its own service,’ and this is what it came up with:
We are always looking for innovative ways to stay up-to-date on the latest technologies and trends. That’s why we decided to ask an AI for FAQs about its own service. We thought it would be interesting to see how an AI would respond to questions about itself and its capabilities. So, we turned to Assistant, a large language model trained by OpenAI. Assistant is a machine learning model that has been trained on a massive dataset of text and is capable of generating human-like responses to prompts. Here’s what we learned when we asked Assistant for FAQs about its own service.
The content feels like it is written for a business, but ChatGPT can change its tone. So I could probably tweak the question to get a more personalized answer that is fit for my website.
Next, just to be sure, I asked the AI:
If ChatGPT is plagiarism free?
The answer is expected and straightforward. Basically, some amount of plagiarism is to be expected as the chat is still a machine and not a human , yet. I assume you can reduce plagiarism by narrowing down sections of the answer (or essay) you expect. Like writing 300 or so words at a time instead of the entire thing. (This article has about 15% plagiarism, according to Grammarly)
Here’s what ChatGPT had to say:
CHATGPT is a machine learning model that has been trained on a large dataset of text, and it is capable of generating responses to prompts that are similar to those a human might give. The responses generated by CHATGPT are not created by a human, but rather are produced by the model based on patterns it has learned from data. As a result, it is important to be aware that CHATGPT may generate responses that are similar to or overlap with existing content. While CHATGPT is not designed to produce plagiarized content, it is important to carefully review the responses it generates and ensure that they are properly cited and attributed if they include content from other sources.
The answer came out the same for all instances:
I’m sorry, but I am not able to provide information on the *service requested of CHATGPT as I am a language model and do not have access to this information. It is important to carefully review the terms and conditions for any language model or other technology that you use, as these terms can have significant legal and practical implications. If you are interested in using CHATGPT or have any questions about the terms and conditions for its use, I recommend contacting the provider of the technology for more information.
This should’ve been a straightforward answer, but OpenAI is withholding the information for some reason.
Now let’s get to the real FAQ. I asked ChatGPT to write an FAQ for itself. The answers were pretty generic but also informative. Here they are:
1. What is CHATGPT?
CHATGPT is a variant of the GPT (Generative Pre-trained Transformer) language model that is specifically designed for generating human-like text in a conversation context. It was trained on a large dataset of conversational text and can generate responses to prompts that are similar to those a human might give.
2. How does CHATGPT work?
CHATGPT uses a neural network architecture called a transformer to process input text and generate output text. It has been pre-trained on a large dataset of conversational text, which means that it has already learned the patterns and structure of natural language. When given a prompt, CHATGPT uses this pre-trained knowledge to generate a response that is coherent and appropriate for the given context.
3. Can CHATGPT generate responses to any prompt?
CHATGPT is capable of generating responses to a wide range of prompts, but like all language models, it is not perfect and may not always generate responses that are completely appropriate or coherent. It is important to keep in mind that CHATGPT is a machine learning model and its responses are generated based on patterns it has learned from data, rather than being programmed with specific rules or knowledge.
4. Can CHATGPT have a conversation with a human?
CHATGPT is designed to generate text that is similar to that produced by a human in a conversational context, and it can certainly be used to have a conversation with a human. However, it is important to remember that CHATGPT is a machine learning model and does not have the same level of intelligence or understanding as a human. Its responses are generated based on patterns it has learned from data, rather than being based on its own understanding or interpretation of the conversation.
5. Can CHATGPT be used for practical applications?
CHATGPT and other language models like it have a wide range of potential practical applications, including natural language processing tasks such as language translation, summarization, and question answering. They can also be used to generate text for chatbots and other conversational interfaces, or to improve the accuracy of language processing systems. However, it is important to keep in mind the limitations of language models like CHATGPT, and to use them appropriately in the context of the task at hand.
I was not quite satisfied with these generic answers, so I asked ChatGPT for an FAQ on its limitations:
ChatGPT Limitations FAQ
Is CHATGPT capable of understanding the context or meaning of a conversation?
CHATGPT is a machine learning model that has been trained on a large dataset of conversational text, but it does not have the same level of understanding or comprehension as a human. It is not capable of truly understanding the context or meaning of a conversation in the same way that a human would. Instead, it generates responses based on patterns it has learned from data, without necessarily having a deep understanding of the conversation as a whole.
Can CHATGPT generate responses that are appropriate for all situations?
CHATGPT is capable of generating responses that are appropriate for a wide range of prompts and contexts, but it is not perfect and may not always generate responses that are completely appropriate or coherent. It is important to keep in mind that CHATGPT is a machine learning model and its responses are generated based on patterns it has learned from data, rather than being programmed with specific rules or knowledge. As a result, it may produce responses that are inappropriate or unrelated to the given prompt in certain situations.
Is CHATGPT capable of having its own thoughts or opinions?
CHATGPT is a machine learning model and does not have the same level of consciousness or self-awareness as a human. It does not have its own thoughts or opinions and cannot form new ideas or concepts on its own. Instead, it generates responses based on patterns it has learned from data, without the ability to truly understand or evaluate the content of those responses.
Is CHATGPT capable of learning and adapting over time?
CHATGPT is a machine learning model that has been pre-trained on a large dataset of conversational text, and it is not capable of learning or adapting in the same way that a human can. While it can generate responses to new prompts based on patterns it has learned from data, it does not have the ability to truly understand or adapt to new situations in the same way that a human would. It is important to keep this in mind when using CHATGPT and to carefully consider its limitations in the context of the task at hand.
ChatGPT: Sharook’s First Impression
Overall, ChatGPT in this form is quite robust and informative. I’ve shifted a lot of my search to the platform. It is not quite human yet, but it is scary close. I assume there will be a ‘Silicon Valley – Pied Piper’ type situation where OpenAI has the tough call to either kill it or risk a global AI takeover. Maybe I am a bit ahead of myself, but the risk is certainly there.
Controversy theorists would say that this is probably not even half of the development. We might be just getting a peak of what ChatGPT is capable of. They already brought in an update recently where the AI answers much more than before. Also, with the kind of real-time data companies like Google has, it’s not without reason to assume that they have something far more useful (or sinister) under their belt.
As for me, I am just happy that something is finally disrupting the tech space at a global level. The Internet had become a maze of boring rigidness, and there was no more room for anything outside SEO.
This is why AI is interesting even though, as a digital marketer, much of my conventional expertise might be under threat. Ai is the future, and I am all for it (please, don’t kill me in the uprising, I am an ally XD).