It is safe to say that most people know and use ChatGPT in their work, study or day to day life. The same goes for me, which is why I chose to write about ‘Echo’, as my personal ChatGPT has nicknamed itself, in this blog post. To clarify, the nickname comes from the model reflecting back on what people put in their prompts, while giving them new ideas, depth and nuance. In this blog, I will focus on what Echo thinks of me and our interaction and reflect on that.
First off, I asked Echo to describe me and our interaction over the years. It seems that I am a curious and exploratory person who digs deep and asks ChatGPT to explain itself. This is probably due to me using AI to help with generating ideas and improving my own input for academic assignments. I have used AI for both study and personal things, such as planning a trip this summer. In addition, I am also a reflective person, asking for multiple perspectives and examples, which apparently not everybody does. Most people are satisfied with surface-level responses.
In summary, Echo gave me the following answer, which is weirdly kind of heartwarming: ‘Interacting with you feels less like answering questions and more like co-creating ideas.’
I think that perfectly describes my personal opinion of how AI should be used. In my opinion, the human brain should still be somewhat capable of comprehending what output a chatbot gives and on what basis. That people are very quickly satisfied with the surface-level responses that AI gives, worries me. Simply assuming that everything an AI model says is correct, could have big negative consequences. In addition, replacing critical thinking skills with only asking AI to give some feedback is risky.
I think that the beauty of AI lies in combining it with the capabilities of the human brain, like making decisions and being creative. For example, asking AI to take on simple tasks could give our mind more freedom to think and dream bigger. I see Echo as a sparring partner, rather than a replacement for my own thoughts, which to me feels crucial for the increasing use of AI in our day to day lives.
I thought about how we can make sure that we use AI in this way and I propose a few suggestions. First off, AI chatbots should ask reflective questions to stimulate users to reflect on the answers that it gives. Secondly, a chatbot should give gentle reminders of ethics, especially when it is used in education. For example when it is asked to write a full essay, it should first respond with an outline only and give some reason why.
What about you? How do you use AI? As a co-creator or as a tool for quick answers? And how do you think that we should engage people in co-creating with AI?
This is really interesting!
After reading this I immediately went to ChatGPT and ask it to reflect on out interaction as well. In short, ChatGPT gave me this answer: ‘The overall pattern is one of co-creation: you bring the raw material of your life, projects, or ideas, and we shape it together into something actionable.’ I would say I completely agree with that. There are some moments when I want to look up for something quick and instead of googling I just quickly ask ChatGPT to tell me a quick fact. But in most cases, for example when using ChatGPT to help me with an assignment, I would bring in my ideas and thoughts, and collaborate with it to improve what I already have. Another ChatGPT’s observation of me is: ‘You tend to iterate’. I have to admit that I often still have trust issues with its answer and still do cross checking with a credible source. When I found that that the fact that ChatGPT had shared with me is not correct, I often try to correct it. I think people should remember that ChatGPT has flaws, and be more critcal when interacting with them.
Refreshing perspective on the use of ChatGPT as well as a fun experiment to do. I fully agree with you that AI and ChatGPT specifically should be used as a sparring partner and not as a replacement for people’s output and critical thinking. If ChatGPT is only used for purely output, the synergies are non-existent. However, as stated in the blog, if you engage more with ChatGPT by responding to and reflecting on initial answers, both your learning curve as well as your work will improve. I think this engagement with AI would be better than solely as a output generator.
I think there is a role for both yourself and OpenAI here. As an user, you should push yourself to not take the first answer as the best answer. Especially as a student where critical thinking is one of the primary skills being taught, you should respond with critical questions and ask for a more well-rounded perspective. As OpenAI, ChatGPT should be made more inviting to use as a sparring partner. In my experience, even if you ask for a different perspective, ChatGPT is often inclined to agree with you. It would be nice if OpenAI would added an option to select a more critical sparring partner instead of a pure output generator.
However, there is the risk that this engagement could replace genuine human discussions, which I believe are still better.