While enjoying a short break from work I came across an article that I thought was “click bait”, until… I clicked and read it.
The gist of it: “AI can write poems (and training, and articles, etc.) but can’t do math”… Ok, I’ll bite.
Circa January of 1997 I was a sales engineer at National Oilwell Varco (NOV) and we were generating business proposals “by hand”. Using macros and coding, we created a configurator or sorts that generated the written proposal and did the quote pricing - all in MS Word. As the code matured, we connected Excel to the Word application and voila - we had our own organic and integrated CPQ solution. One tool did the “copy”, another the “math”; mission accomplished.
Just because you can, it doesn’t mean you should (I’m still talking business solutions here, stay with me). Microsoft Word was not designed to to math, but “can do some”. Generative AI was not specifically created to do math, and quite frankly, while it can use a broad knowledge base to explain to me how do design a submersible to see the Titanic, I wouldn’t rely on it to tell me what the safe test depth of the vehicle would be. (Too soon?)
“In the ever changing world of AI and this era of misinformation…” - no, I did not ask AI to write this article, but we recognize that current AI models, even the most sophisticated, have their limitations. Like repeating the lede to this paragraph in 90%+ of the write ups folks ask AI to generate.
GenAI is a formidable equalizer and technology that every SMB should integrate to their technical stack. Connecting multiple GenAI applications and “passing” prompts and results from one to the other is NOT a flaw; it is smart architecture and business savvy. Leverage the best technology has to offer now; while it continues to mature into more formidable and complete solutions.
That said -
Don’t ask GenAI to design the next mission to the moon or mars. Perhaps consider using GenAI to automatically transcribe and analyze the dialog between the spacecraft operators.
Don’t ask ML to write poetry a la Bill Shakespeare, or stories like Cervantes. Perhaps use it to analyze patterns.. call me crazy.
Even when you use GenAI to create all the graphics of your next presentation, consider that humans only have one head and for the most part, less than three arms…. Check the work product.
GenAI is not perfect, and perhaps can’t do differential equations today. I got money that in my lifetime it will help us solve some problems that have kicked our hide and imagination.
While not perfect; don’t fall for click bait on either side of the spectrum. Consider carefully its use, and even more how you deploy it.
Don’t use a wrench to drive a nail, even though sometimes it works just fine.
Here is the link to the article I was reading... :)