A chat with and about ChatGPT

Jamie Biesiada

As technology evolves, the question of whether it will render humans obsolete looms larger. Let’s dive into the possible future and see if our jobs are truly at risk.

Full disclosure: I didn’t write that first paragraph. The artificial intelligence bot ChatGPT did, and I think it answers my original question of whether technology will one day replace us all: It won’t, at least not yet. It’s OK, but I asked ChatGPT to write it in my voice; this was probably too specific a request, and specificity is an area in which, I’ve found, the program struggles. I don’t think I’d ever use the phrase “Let’s dive into” in a lead paragraph.

At least ChatGPT is self-aware (computer-aware? algorithm-aware?) enough to know it’s not ready to replace humans. For a recent story, I asked it if it could replace a human travel advisor. Its answer: Not fully.

I’ve been toying around with the tool for a while now. I do look at it as a potential tool, if not for myself, then for others. My biggest takeaway remains that while it’s impressive, it’s not nuanced enough to start taking over jobs (not yet, at least).

In short, I’ll go ahead and call it “pretty OK.”

But I do see some potential uses for travel advisors. Real estate agents, apparently, are already using the service to save time when generating listing descriptions.

For travel advisors who don’t consider themselves writers, ChatGPT might be helpful in populating custom itineraries or proposals. But I say this with caveats. Take, for instance, the response to a request I made to ChatGPT to “write a short description of the Battle of Culloden, its historical significance and what visitors will experience at the battlefield.”

Its answer was pretty OK. It covered the basics of the battle as well as the monuments there and interactive exhibits at the visitor center. I wouldn’t be upset if its response was inserted into an itinerary for a trip to Scotland.

I was less impressed with the result for this prompt: “Write a short description of how Walt Disney World’s Genie and Genie+ services work.”

Its answer was decent, but it references “access to private clubs” for guests who purchase Genie+ — Disney, did I miss that announcement?

That error brings to the forefront the importance of bringing the human touch to whatever response artificial intelligence generates. It’s what AI lacks today, and what clients want when working with advisors.

Its use also raises ethical questions. Its creators have been upfront about the fact that it sometimes generates incorrect answers, or answers that don’t make sense. And it usually doesn’t ask clarifying questions, but instead guesses what the prompt-writer actually wants. The spread of misinformation is an active and ongoing threat in the world, hence the need for fact-checking.

One article I read in the New York Times suggested that a user disclose their use of ChatGPT. Something to consider.

I’ll leave you with my new bio. Disclosure: It was written by ChatGPT. And it’s, well, pretty OK.

“Meet Jamie Biesiada, a travel industry professional with a love for adventure and a knack for finding the best deals. When she’s not jet-setting around the world, you can find her sipping on a margarita and planning her next getaway. Her secret to success? Never taking herself too seriously and always being willing to try new things. Just don’t ask her to choose between the beach and the mountains, it’s a lose-lose situation.

“She’s been known to have a sharp wit, a taste for good food and a love for travel trends. Jamie is a force to be reckoned with in the travel industry. She’s always on the lookout for the next big thing, so watch out world, Jamie’s coming for you.”

Source: Read Full Article