Written by Mark Evilsizor
From his column Tech
Haga clic aquí para ver esto en español
Within the first five days of its November 2022 release, ChatGPT, from the company OpenAI, signed up over a million users. Its ability to create, with little prompting, human-like dialog and essays has set the technology and academic worlds all of a flutter, and it is poised to disrupt writing in every field. Let’s take a look at this amazing technology, how it is being used now, and how it may be helpful or harmful in the future.
If you allow your text messaging app to predict words or use prompted sentence completions in Word or Gmail, then you have experienced an inkling of what is now possible with generative AI. General Artificial Intelligence (AI) is the notion of a machine starting out with only the ability to understand and learn. Then, like a child, that machine, through experiences and training, becomes capable of anything—playing chess, drafting legal documents, performing brain surgery, or piloting rockets to other planets. More than this, that machine interacts in a humanlike manner while performing these tasks.
Although it sounds similar, Generative AI is not the same. It is a software that has been fed miles of text, in ChatGPT’s case, more than one million feet of bookshelf space of textual content from the internet. Based on this massive trove of data, this software is able to recognize patterns and qualities of various styles of writing which enables it to interact with humans, and generate content when prompted.
For example, when Thomas H. Ptacek gave ChatGPT the following prompt: “Write a biblical verse in the style of the King James Bible explaining how to remove a peanut butter sandwich from a VCR,” its response began like this:
- And it came to pass that a man was troubled by a peanut butter sandwich, for it had been placed within his VCR, and he knew not how to remove it.
- And he cried out to the Lord, saying, “Oh, Lord, how can I remove this sandwich from my VCR, for it is stuck fast and will not budge.”
People are just starting to scratch the surface of how this system can be tapped. Realtors are using it to write blurbs for houses they are selling, so what used to take 20 minutes now takes 5 seconds. Feed it the facts and it writes enticing prose in response. A tech news company was recently faulted for using ChatGPT to write articles after the authoritative-sounding pieces were discovered to be factually incorrect. Another company, specializing in punchy lists and personality quizzes, is gearing up to generate a large number of articles automatically. This has their paid writers in an uproar. Of course, students have discovered it and have tried to submit homework essays written by ChatGPT. You can even induce it to include grammatical errors, so the work doesn’t appear too perfect. It has become prevalent enough that some professors are requiring first drafts of essays to be handwritten in class.
When we look back at the transitions from slide rule to calculator to math software, it is usually with a grin at how advances in such tools made our lives so much easier. But somehow, with words instead of numbers, it feels different.
Some clergy friends have shared how they love ministering to the people in their congregation, encouraging in times of despair, multiplying their own joy by sharing in life celebrations, and listening empathetically when others need support. But they have also told me they anxiously struggle to create fresh, meaningful sermons every week. Should they consider ChatGPT as a helpful tool? What impact would that have on the power of the Holy Spirit to guide the message God has for a pastor and their congregation?
I asked ChatGPT to write a modern version of the parable of the Good Samaritan. With an initial prompt and three iterative refinements, it created a story set in Vietnam, culminating in this paragraph:
- The wounded soldier, who was in shock and pain, asked the farmer, “Why would you risk your own safety to help me? I am your enemy.” The farmer simply replied, “In this moment, I see only a fellow human being in need. I may not have much, but I have the ability to help and make a difference in someone’s life. That’s all that matters.”
It may be we will decide that in some situations the cost of allowing a computer to express our words comes at too great a cost. We will have to decide in what areas and to what extent we will invite this type of technology into our lives. What is lost if students generate prompts and don’t work through the instructive struggle of expressing the thoughts and feelings they have inside?
One person wanted to connect a computer to generative AI to serve as their lawyer in court. The judge put a stop to it, but one can imagine how such software can be used in the justice system.
There will be fascinating discussions in the months and years ahead as we work through defining how intrusive and helpful a role we will allow computers to have in this most human endeavor of expressing ourselves. And, in case you were wondering how the fight between the peanut butter sandwich and the VCR ended up, it went like this
- And the Lord said, “Verily I say unto thee, seek not to put thy peanut butter sandwiches in thy VCR, for it is not a suitable place for such things. Rather, keep thy sandwiches in thy refrigerator on thy plate, where they belong.”
Mark Evilsizor has worked in Information Technology for more than 25 years. He currently serves as head of IT for the Linda Hall Library in Kansas City, Mo. Opinions expressed are his own.
The following written pieces are the product of artificial intelligence created by ChatGPT based on prompts by author Mark Evilsizor.
Appendix A – The Brief Sermon in Full
Appendix B – Retelling the Parable
Appendix C – Micah 6:8