Editing a ChatGPT Story

So I've never been into the Artificial Intelligence (AI) scene. My interactions inlcude Siri and products like iRobot. But this is not the same type of AI we as humanity are starting to interact with on the daily. Siri and iRobot are the type that are meant to follow orders, prompts, or a set of instructions and deliver an outcome. They are not meant to be subjective.

ChatGPT, full name Chat Generative Pre-trained Transformer, originated from a company in San Francisco called OpenAI, with one of it's founders being none other than Elon Musk, a big name in the Silcon Valley. While OpenAI Inc. is a non-profit Organization, it has a seperate entity called OpenAI Limited Partnership (OpenAI LP) that is for profit. For now, it seems that ChatGPT is free, but major news outlets have covered that Microsoft has invested billions into the new technology, so this freedom may not last for long.

In fact, the AI might not last as CNBC reported on Wednesday, April 4, 2023 around noon via a YouTube poll that "Italy became the first Western country to ban ChatGPT." Yet even with our worries, the rise of AI has taken quite a lot of the world's interest with star YouTubers Rhett and Link of Good Mythical Morning and Kian and JC posting videos interacting with AI. Vogue-runway on Instagram posted on Saturday, April 8th, of the first New York AI fashion week (April 20-21) being held at the Soho's Spring Studios, quoting "the event is making the case for AI as a tool for fashion design." Iceland is reportedly using a more advanced version of ChatGPT (ChatGPT 4) to preserve its language and it's been intergated into educational companies like Duolingo and Khan Academy.

So what is it? Those were my thoughts as I had never interacted with ChatGPT, so in order to fully comprehend its capabilities and limitations, I signed up - warily.

The reason why this disclaimer pops up may have to do with the fact that like any good research, ChatGPT is not the first model and certainly won't be the last working one either. Before ChatGPT came several models were finetuned into its final iteration, but even then a succinct model called InstructGPT seems to be ready at the helm. This InstructGPT is imagined to take instructions and perfom based on the vague guidelines set forth by the inquirer. ChatGPT however is far different in that it provides lies and garrulous information that can distract from the inquirer's main question. With this in mind, lets analyze my first interaction with ChatGPT, keeping its limitations in mind:

ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers.

My first impression is that the chatbot found the need for a numbered list as my prompt was "tell me," implying a variety of ways to release the information. I had assumed it would use a paragraph. The second is the importance of the list, one obviously being listed before seven, but does creativity trump honesty? I think not, especially in the world of journalism and academic press. This makes me believe the chatbot scanned the internet and found the most popular wesbites using SEO data to deliver the fastest information possible.

Is it the most reliable? No, as there are no citations or any reference as to where this may have come from which is why I can see from an educational standpoint, this being a setback. Just like the days where Wikipedia was considered flawed and therefore shunned for it's open interactive interface, we must find a way to incorporate these new pieces of information into the standard we trust. ChatGPT is like Wiki in that it's a secondary source. It's pulling information from somewhere and without the necessary tools to find the source and then deem it educational, it can't be trusted.

But can a chatbot imagine? Can it create like a writer? Well, I tested that out too:

ChatGPT follows a standard model of storytelling here. It's oversimplified but to the point with none of that verbose language we were warned about in its limitations:

The model is often excessively verbose and overuses certain phrases, such as restating that it’s a language model trained by OpenAI. These issues arise from biases in the training data (trainers prefer longer answers that look more comprehensive) and well-known over-optimization issues.1

First, the use of "once upon a time" grates me. The phrase itself has no setting, no way for the reader to settle into the scene, and it's contradictory. Once does not not happen within time as time is a continuation of something. Time is the essence by which we measure the present, the past, and future, and without knowing what time, what setting, what era it is, we can never truly know if it happened once.

Now in fairytales, it doesn't matter because it's fictional, but this is a parable - a fictional story meant to teach a lesson. Parables have only one criteria, they must have a universal lesson, and this stories universe runs on some point of time - so point off for the lack of "when" setting. We then move to the general settings - who (Mei), what (young girl), when (once upon a time), where (small village in moutains), and why (loves to explore). The scene quickly adds climax (bird with broken wing), action (taking care of the bird), the falling action (setting it free), and end (lesson of kindness and compassion.)

Overall, it's a simple story and sometimes those are the best, but will it overtake a writer? Will one day our jobs as storytellers, as documenters and narrators, copywriters and editors be over taken by this or some other model? Well, on March 17, 2023, a few authors reported in this article GPTs are GPTs that we may see as much as "80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of GPTs, while around 19% of workers may see at least 50% of their tasks impacted."

It is without a doubt that the introduction of AI will have socio-economic impacts worldwide, but it is a challenge that I think humanity can face if we only remember one thing: AI was made for us and by us. If we lose our sense of control and start to fear the technology we are making, that is when we have lost our understanding of humanity for we are a people that have never been afraid to overcome another challenge, to race to the future, and to better our world. I am interested to see where this new strain of artificial intelligence will take us.