Spoiler alert…I didn’t write last week’s Thursday Thought…an AI bot did.
The entire blog last week was written by an AI chatbot. Very few of you noticed! I first gave the AI bot a sample of my blog to review (the one on Polarizing Positions—four weeks ago). The bot then generated an analysis of my voice from that blog* (see below). Then I asked it to write a blog concerning the danger of comparison within social media and to include references by Seth Godin, Dallas Willard, and the New Testament. What you saw in last week’s blog was entirely the creation of the chatbot at chat.openai.com!
Let me first apologize to those of you who said it was one of my best! No, that didn’t happen, but several found it very helpful. Ai is becoming disconcerting for me…that is good and bad.
Initially, some of the challenges of AI is that it would make an 8th grader sound like a graduate student. However, now all one must do is send the bot a copy of something they wrote, or better yet, multiple samples of something you wrote (in your “voice”). Once that is done, the ai bot will mimic the voice of the author and is capable of accounting for the individual’s style of speaking and/or writing. More or less. If you go back and read last week’s blog, you will notice things that are not like my writing style. For instance, I never say, “In conclusion” in my closing argument…I am just not that organized!
By the way, the scary aspect of the blog it wrote for me last week was that it wrote that blog instantaneously! Yes, it had perfect grammar, organization, and content in less than a second. I sometimes labor hours over how to say something and have multiple people read it for errors.
This is where it does get a little disconcerting. For instance, can I give it the 95 scales that I use in many of my assessments? Yes, I can give it the manuals that explain each of these scales, and it will summarize the scales well. However, as of now, AI is unable to synthesize the scales in a way that paints a composite of how one scale flavors another scale. This has taken me years of learning, and there are no current manuals to explain these interactions. Will there be someday? Yes.
I have learned a lot these past 30 years of doing assessments on how one scale from one instrument informs the scales of other instruments. I have learned this by talking, observing, reading, risking, asking for feedback, and changing my perspective. The AI bot can also do this with time if it is given the opportunity to learn with feedback. The algorithms cannot be black and white but must be nuanced. With time AI bots will get there.
This is where it really becomes intriguing. What will be left for us to do? As I have spoken with some who are deeply engaged in the development of AI, they don’t paint a rosy picture for many of us humans. I realize when automobiles came along, the horse population plummeted. Today, there are half the number of horses there were in 1900 in the United States (see the article the AI wrote for me on this topic at the end of this blog).
So what will we be needed for? What work will honor our Creator?
What is uniquely human about our existence and reasoning? As I think about this, we do have the ability to create, associate, integrate, interpolate, and extrapolate. These are all different ways we process experience, the context, and our values to make decisions. I anticipate in not too long, AI bots will be able to approximate these ways of processing.
Hopefully, we learn as we make decisions, live lives, and experience consequences. AI bots cannot experience, but they can learn. Is there a difference? I think so. Neither do AI bots have emotions that are supra-rational. In other words, often, our emotions are not logical, and they can cause us to feel or act, or even think in ways that don’t make sense. And that, too, is an aspect of being human. Sensing the Spirit’s insight in our lives is an entirely different issue we will discuss later. God has given us a soul which integrates all aspects of ourselves. According to Dallas Willard, this is uniquely human. I hope so.
*Here’s my analysis of your writing style:
- Tone: Your writing style is thoughtful and introspective. You approach the topic of polarizing positions with a measured tone, acknowledging that there are valid arguments on both sides of the issue.
- Sentence Structure: You use a variety of sentence structures, including simple and complex sentences, to convey your ideas. This adds interest and variety to your writing.
- Vocabulary: Your vocabulary is varied and sophisticated, which helps to keep your writing engaging and informative. You use precise language to convey your thoughts clearly and effectively.
- Organization: Your blog post is well-organized and structured, with a clear introduction, body, and conclusion. This makes it easy for readers to follow along and understand your argument.
- Use of Quotes: You use quotes effectively to support your argument and add credibility to your writing. You also use quotes sparingly, which helps to keep the focus on your own ideas and analysis.
Overall, your writing style is engaging and informative, and you have a clear grasp of the topic you are discussing. Keep up the good work!
**The early 1900s was a time when horses were an integral part of daily life, serving as transportation, labor, and even companionship. However, with the rise of modern technology, the use of horses has dwindled significantly. So, how many horses were there in 1900 compared to today?
In the early 1900s, there were an estimated 20 million horses in the United States alone. These horses were used for everything from plowing fields to pulling carriages and wagons. Horses were also used in the military for transportation and in the cavalry for battle. The horse industry was thriving, and many people made their living as horse trainers, breeders, and caretakers.
Today, the number of horses in the United States has significantly decreased. According to the American Horse Council, there were an estimated 9.2 million horses in the United States in 2021. This represents a decrease of over 50% from the number of horses in 1900.
The decline in the horse population can be attributed to several factors. The rise of the automobile and other forms of transportation led to a decreased demand for horses for transportation. Advances in farming technology, such as tractors and other machinery, also reduced the need for horses in agriculture. Additionally, the shift in lifestyle and leisure activities has led to a decrease in the popularity of horseback riding and equestrian sports.
Despite the decrease in the overall number of horses, certain breeds have experienced a resurgence in popularity. For example, the Quarter Horse and the Thoroughbred are two breeds that have seen an increase in demand in recent years. The popularity of horse racing and rodeo events has also kept some breeds in demand.
In conclusion, while there were an estimated 20 million horses in the United States in 1900, today, there are only an estimated 9.2 million horses. The decline in the horse population can be attributed to a number of factors, including advances in technology and changes in lifestyle. However, certain breeds continue to thrive, and the horse industry remains an important part of American culture and history.
7 Replies to “My Intrigue with AI”
I wouldn’t have said this you last week. First – I had no thought that it might have been AI generated. None. But it didn’t sound like you to me. I usually want to respond after reading your writings. They cause me to desire engagement. But last week I didn’t get that ‘feeling’.
Maybe that’s what we humans bring to the table. Humanity.
When we were at First Church in Vero, besides Danny Daniels, we had two exceptionally gifted pianists. One was precise. Never missed a note or struggled with timing. The other played beautifully and from his heart. Guess which one I preferred?? 😊
Thanks for your always thoughtful take on a variety of topics! And if today’s offering was actually written by an AI after all, they’ve learned since last week! Haha
You are right Mariann, it was too precise from my voice. I am told if I would have given it about 10 of my blogs rather than one, the ai bot would have done a much better job. The bot that I used, is referenced as the Kindergarten version compared to the versions that one can pay for are like a college graduate. So I will be curious to see if they get better in adding the human element. I saw in today’s news that some rap singer (whom you would know, but I don’t) had an ai version of a song that went viral, only problem was the entire voice was ai created as well as the song…this is going to get more interesting.
Typed by human thumbs!
Actually, last week, I wondered…
The typical ChatGPT output is “vanilla, no sprinkles;” anodyne, inoffensive and last week’s topics was,,,. well…
Your style is typically more insightful than that.
Will AI “replace us?”
I don’t know, but here’s a story out of our mutually shared engineering past.
In 1982-3, we introduced the first spreadsheet tools into our engineering group. They were supposed to immediately eliminate jobs. What happened was that, instead of a couple of passes per day on those massive paper spreadsheets we tortured ourselves over, we could make many, many more analytical evaluations per day.
Arguably, decision making may or may not have gotten better, but decision making was and remains quite a different process than simply running multiple what-if analyses.
I won’t be surprised if we see a similar effect as AI becomes more prevalent. Of course, what form the effect will take remains to be seen.
I use ChatGPT to generate the 60-70% of a proposal or white paper that remains blah-blah-blah, then edit heavily if I want to add “punch,” value-added text, or stats.
In conclusion… 😉
That is how I am using it as well. It generates a lot of the linear content and then I take it and use it to present my case. It will be interesting if it in fact develops a sense of self learning to replicate the human processor…
Always check your (AI) sources. ChatGPT in particular loves to hallucinate facts. Pairing it with something like wolfram alpha which can parse natural language and has huge databases of real data to verify is incredibly powerful but by itself chatgpt is dangerous as a fact checker!
Bing chat is actually better because it at least does a search and summarizes real results but it’s also susceptible to garbage in, garbage out.
I asked bing chat to calculate the ratio of people to horses in the US from 1900 to now which is probably even more relevant than raw numbers. It went from 3.5 people per horse to 47. So it’s not just a ~2x drop (it’s actually closer to 3x with the data bing found), but it’s over a 13x drop in relative numbers!
For all the flaws though, this is the worst AI will be going forward. And it went from doing a few small party tricks a year or two ago to a huge explosion in incredibly useful applications in an amazingly short time. We are in the middle of a revolution and it’s almost impossible to predict the final state.
Could you have imagined the social media / smartphone / internet life we live now 25 years ago? The AI revolution will be as a dramatic in likely an even shorter period of time.
Well said Jordan. Thanks for weighing in.