Remember the early days of Microsoft Word, when WordArt and Comic Sans font were the epitome of cool? You'd spend hours tweaking the colors, shadows, and shapes of your text, thinking you were creating a masterpiece. But let's be real, looking back at those documents now is like opening a time capsule of cringe.
Well, buckle up, because the world of AI is experiencing its own WordArt phase, and it's all thanks to poor Large Language Models (LLMs). These AI systems are the Comic Sans of the tech world – they're everywhere, and people can't seem to get enough of them, even though they're not always the most aesthetically pleasing or functional choice.
Just like how everyone and their grandma started using WordArt and Comic Sans when Word first introduced them, people are now jumping on the AI bandwagon, using poor LLMs left and right. And yes, in the beginning, it might seem interesting and novel, like adding a dash of pizzazz to your school project or work presentation. But trust me, as time passes, the novelty wears off, and you're left with a bunch of AI-generated content that's about as appealing as a Comic Sans font on a professional resume.
However, just like there were always those few individuals who knew how to make WordArt look great (you know, the ones who actually read the manual and experimented with the settings), there is hope for those who know how to prompt LLMs effectively. By understanding the intricacies of these AI systems and crafting well-thought-out prompts, you can actually generate content that doesn't make people want to gouge their eyes out.
But here's the thing: I'm tired of receiving emails (not from my students) that are clearly written by someone who just put summaries through an AI or used a prompt to write an email asking for a job position. I know what AI-generated text looks like, and it's about as subtle as a dragon in a tea party. Speaking of which, I recently received an email that could have been straight out of Game of Thrones. Here's a snippet:
"Esteemed Dr. Rocha,
I am writing to inquire about the possibility of securing a position as a research associate within your esteemed laboratory, which focuses on the intricacies and nuances of dental materials.
As a recent graduate with a Master's degree in Materials Science and Engineering, I possess a keen interest in the field of dentistry and a fervent desire to contribute to the advancement of dental materials research.
My academic background, coupled with my innate curiosity and aptitude for scientific inquiry, has prepared me to undertake the challenges and responsibilities associated with this role. I am confident that my skills and knowledge would be an asset to your team, and I am eager to delve into the depths of this fascinating field under your tutelage.
I implore you to consider my application and grant me the opportunity to further discuss my qualifications in person. I assure you that I am as dedicated to this pursuit as the Night's Watch is to guarding the Wall.
Yours sincerely,
###Applicant Name##"
I mean, come on. "Innate curiosity and aptitude for scientific inquiry"? "Delve into the depths of this fascinating field"? "As dedicated to this pursuit as the Night's Watch is to guarding the Wall"? It's like they fed a thesaurus and a Game of Thrones script to an AI and hoped for the best.
So, the next time you're tempted to jump on the AI bandwagon and start using poor LLMs willy-nilly, remember the lessons we learned from the WordArt era. Take the time to learn how to prompt these systems properly, and you might just create something that stands the test of time – or at least doesn't end up in a "Worst AI-Generated Content" compilation a decade from now.
In the meantime, let's all collectively agree to leave Comic Sans, poorly-used LLMs, and Westeros references in the past, where they belong. Your eyes (and your inbox) will thank you.
If you are interested in learning more about AI and apply to your daily routine. Let me know! I can help you.