“ChatGPT stands as a powerful and versatile language model, offering users assistance and information on a myriad of topics. However, beneath its seemingly benign surface lies a landscape fraught with potential pitfalls and risks that demand our attention. As we embrace the capabilities of ChatGPT, it becomes imperative to explore the shadows cast by this sophisticated AI tool.”
I did not write that opening paragraph, but I did ask ChatGPT to write the beginning of a newspaper column about the dangers of itself.
Artificial intelligence arouses fear, wonder, joy and animosity — it’s captivating in its scope, bewildering in its capabilities. Pew Research states that as of 2022, 19% of American workers were in jobs that could be replaced or assisted by AI.
Over the past year, I’ve included many stories about artificial intelligence in the Daily Journal. While that may partly be due to my own interest in the field, AI has also become impossible to ignore. Media attention on the proliferation of AI is nonstop, itself an ever-changing reaction to the advances of the technology each week.
I am a writer — this technology is somewhat of a threat to the career path I’ve chosen. But it’s also so wildly entertaining, I can’t help but play around and be entertained with it.
As fun as ChatGPT is, it’s just the start of AI. ChatGPT is merely a language model that gathers information and outputs the information into text. Its database of information is billions of pieces of information from the internet, social media, novels and even human conversations. ChatGPT doesn’t possess any signs of higher intelligence, but its rudimentary functions are still fascinating to toy with.
ChatGPT can answer millions of questions, write essays in the style of a first-grader or Shakespeare and be a repository for loads of information. Anyone can use it, even two New York lawyers who submitted bogus case briefs with the technology. School systems around the country are having to adapt to how they handle cases of plagiarism due to AI.
Wired ran a compelling story last year about a ChatGPT user who created a story generator that inserts his family into the world of the children’s show Bluey. Anyone can generate personalized stories starring their children with their favorite characters. It’s a light-hearted story that also creates more questions than answers regarding intellectual property law.
Sora, announced last week by OpenAI, may be the most fascinating AI development yet. It’s not available yet, but Sora is a high-quality text-to-video generator. Give Sora a prompt and it will deliver exactly what you asked for.
Prompt: “A movie trailer featuring the adventures of a 30 year old space man wearing a red wool knitted motorcycle helmet, blue sky, salt desert, cinematic style, shot on 35 mm film, vivid colors.” The accompanying video posted on X (formerly Twitter) matches the prompt so well you would think you were sitting in a movie theater.
What before required a team of professionals and a Hollywood budget of millions can now be accomplished in the span of about ten seconds of writing. Other videos of the Sora technology on X are wildly entertaining; I would encourage anyone to check them out.
It’s not all rosy with the future of AI. The potential for misinformation, lawsuits and deepfake videos is deeply troubling with no real answers in sight. The Washington Post profiled a Google engineer, responsible for interacting with burgeoning AI technology, who believes that LaMDA (Langugae Model for Dialogue Applications) is functionally real and sentient.
Part of what drove my decision to write this column was to stamp my own neophyte knowledge down at this point in time; I feel that even maybe ten years from now, AI could have so many applications in daily life that my novel understanding presented at this time would already seem obsolete. It would be like reading a news article of someone’s inability to understand how to use a computer fifty years ago. There are so many questions surrounding the field of AI, but one thing is crystal clear — The genie is out of the bottle and AI is here to stay. We are all venturing into uncharted territory together.
As we delve into the dangers associated with ChatGPT, it is essential to strike a balance between harnessing its capabilities and acknowledging its limitations. The road ahead involves not only refining the technology itself but also establishing a framework of responsible usage and ethical considerations.
I didn’t write the preceding paragraph. That is both horrifying and amazing simultaneously.