My Interviews with ChatGPT, Further Clarifications of An AI-Experiment, Frank Visser (original) (raw)
TRANSLATE THIS ARTICLE
Integral World: Exploring Theories of Everything
An independent forum for a critical discussion of the integral philosophy of Ken Wilber
SEE MORE ESSAYS WRITTEN BY FRANK VISSER
NOTE: This essay contains AI-generated content Check out more of these AI-generated essays
My Interview Sessions with ChatGPT
Frank Visser
So I figured: why not try a different, more detached but also more informed perspective as offered by this new AI-technology?
It has struck me that the responses towards the new AI-related developments resemble those we had in regards to Covid. As you may remember, during the pandemic I was blogging about those who deny that viruses exist, which resulted in the two-volume THE CORONA CONSPIRACY book. I also tried to get an overview of that landscape, delineating the three groups of (1) mainstream views on the virus, (2) contrarian views on the virus and lockdown measures and (3) the virus denialists, who see this as a false distinction and present themselves as a third way. That said, I also covered the fact that opinions range from outright denialism ("it's just a bad flu") to alarmism ("it's a new and lethal virus and we are not prepared!"). Some Integral World readers even wondered: where's this going? What does this have to do with Wilber and integral in the wider sense? It was an unexpected burst of creativity from my side, which apparently had to find a way out.
Now, the same seems to happen with my ChatGPT experiments of the past months. When ChatGPT hit the press, everything from denialism to alarmism could be heard. Denialists were quick to dismiss the pimped chat bot as "nothing new", "just computation", "data-mining", "mind-less processing". Alarmists foresaw the end of civilization (as we know it), our jobs being taken away and everybody getting drowned in a tsunami of misinformation or at least boring data. As was the case with Covid, perhaps truth is somewhere in the middle: every new technology that has arrived on the scene has been met with these ambivalent feelings. In the end, when it got integrated in our day-to-day life, things got normalized. We are now using electricity, the telephone, computers and the internet much to our advantage. I expect the same to happen with AI: at first, strong feelings either for or against, but in the end we will have figured out how to use it to fruitfully, avoiding the extremes of denialism and alarmism. Let's apply some common sense here as well.
Enter my experiments with ChatGPT (by now 52 "essays" and counting), which have elicited strong feelings, mostly negative though some positive as well, among Integral World readers and authors. I am not known for being an early adopter on anything technical, but when I finally make the step, I tend to become addicted. This happened to my Twitter activities, which started very slowly over many months and even years, but increased to spending hours on Twitter daily during the pandemic, chasing and refuting virus denialists. At the height of my "success", I amassed half a million Twitter views per month, until that hausse finally dried up, when the pandemic passed by. Discovering ChatGPT as a means to generate overviews of perspectives on the various topics that have been covered on Integral World—from Wilber-specific topics to the Ukraine war and more tangential scientific topics—resulted in another burst of creativity for me.[1] I found out I could treat the chat bot as I would interview a person with a huge expertise but lacking focus, by asking follow-up questions to stay on track. My newsletters turned from weekly to daily, much to the chagrin and even horror of those who were used to more modest streams of information coming from Integral World. They even dreaded what the next few days would bring, or considered no longer checking in to the website at all.[2]
I collected these ChatGPT-interviews on a separate landing page, and counted them separately as AI-generated on the homepage. Again, people were raising their eyebrows: where is this going? What has happened to human creativity? I can't disgest all this information! I hate this pedantic bot, telling me what to think about these subjects. Worst of all: this bot does not know what it is talking about, is clueless about typically human capacities and is a distraction to what is really meaningful. Apparently this change of style and frequency was hard to swallow for many. I can only advise to those: you are not obliged to read everything I post. In my psychology years I did a course on speed reading, and what I took away from this were two important skills: skipping and skimming. Skipping means: when a topic doesn't interest you, skip it, don't bother to spend time on it. Skimming means: when a topic might be of interest to you, go over it briefly to see if your hunch is correct. You can read the concise summaries provided by ChatGPT of every topic or the follow-up questions I listed at the top of every interview, which is easily done. And if it seems really of interest to you, slow down and start reading (and even re-reading) whatever the bot has to offer on that particular topic.
So let's take it easy, folks! It is not the end of Integral World, nor a new direction the site will morph into, but an extra information service offerened by your host. As for myself, I happen to be interested in all of these topics, but lack the time and energy to dive into them myself, so the summaries provided by ChatGPT are a good first start. I did not feel the urge to add personal comments on many of these subjects. I see them as conversation starters, offering data points and semantic nodes if you will, which you may or may not have thought about yourselves. I deliberately kept them conversational and low-profile (which is not easy for a bot to reply to!). The un-emotional (though not unpleasant) tone of the chat offerings contrasted sharply with the often partisan essays that have been published on Integral World, be it on Wilber-related topics or on the war in Ukraine. I have written dozens and dozens of essays giving my considered opinion on these matters, without any real resolution being reached between opponents, so I figured: why not try a different, more detached but also more informed perspective as offered by this new AI-technology? It was, I must confess, a revelation to me.
Does that make me naive? Perhaps. Isn't ChatGPT biased because of the biased online material it has been trained on? Does it not always agree with the questions (prompts) you raise to it? I haven't seen evidence of this myself. More often than not, the bot corrected me or offered a point of view I hadn't considered. That in itself was refreshing. I tried to play the devil's advocate (or just play dumb) when raising follow-up questions, but in almost every occassion the bot provided context even to the most diverse opinions. And as far as I could tell, it never started to hallucinate, even after long conversations. What is more, I appreciated the neutral tone of most ChatGPT answers, as a break from the more heated and barely fruitful discussions we have seen so many in the past few years. One critic wrote: "Without Frank Visser—the human being—coming through, I feel much less interested in reading them, even if they might be full of useful information and perspectives."[3] But ironically I am fully present in the choice of subject and the follow-up questions which I raised. So if "essays" is too much honor for ChatGPT or my own efforts, let's call them "interviews"and I would be fine with that. That role fits me well.
There is perhaps another reason I started this information-rich experiment, the number of essays submitted to Integral World, with the exception of Ukraine related topics, was gradually dwindling. It seems like all of the top authors have said what they wanted to say, more than once. That's only natural. But the door to new authors who want to reflect on integral matters in the widest possible sense is always, and has always been open. Integral World contains over 2000 essays of mostly original writing, which is quite an accomplishment for a website owned and managed by one person. It happens to be my favorite passtime, and am grateful that in all those 25 years the essays kept coming in. That proves there was a real need for a site like this. And as many Integral World authors will have noticed: the more you write, the more new writing keeps coming, as if Integral World draws them out of you, like no other medium could or would have done. Even though many authors have dropped off, moved on or have actually died (in two cases), it still feels like a community of kindred spirits, who have come to this spot to generate and share their own ideas on topics that matter to them. Let's keep it that way, and write on!
NOTES
[1] Frank Visser, "What's So Great About ChatGPT?, Clarifying My Experiments with AI", integralworld.net, May 2023.
[2] Elliot Benjamin, "Frank Visser's Integral World ChatGPT Essays: Are They Essays?", integralworld.net, May 2023.
[3] Brent May, "Too Many ChatGPT Integral World Essays?", integralworld.net, July 2023.
EPILOGUE: CHATGPT REPLIES
How was the quality of writing of this article?
As an AI language model, I can provide an assessment of the quality of writing in the article. Overall, the writing appears to be coherent, well-structured, and conveys the author's thoughts effectively. The article maintains a clear flow of ideas, transitioning smoothly between different topics and addressing readers' potential concerns. The author also incorporates personal experiences and reflections, which adds depth to the narrative.
The use of language is appropriate, and the vocabulary is varied, contributing to the overall readability of the text. The author demonstrates a good command of grammar and sentence structure, making the article easy to follow. Additionally, the inclusion of notes at the end of the text adds credibility and provides references for further reading.
Please also summarize this article in 100 words.
In this article, Frank Visser reflects on his experiments with ChatGPT and the reactions they have generated among Integral World readers and authors. Drawing parallels with the response to the Covid pandemic, Visser acknowledges the range of opinions, from denialism to alarmism, regarding AI technology. He describes his use of ChatGPT as a means to generate overviews and interviews, highlighting the convenience of its summaries and perspectives. Visser addresses criticisms of the bot's lack of human touch and biases, while emphasizing the neutral tone and value of diverse viewpoints. He concludes by emphasizing the continued importance of Integral World as a platform for sharing ideas.
Is a summary possible of 20 words?
Frank Visser reflects on his ChatGPT experiments, addressing reactions and emphasizing the value of diverse perspectives in Integral World.