Want Better Recommendations? Try Chat GPT
Yes, it’s another post about AI generated content. How original – everybody’s doing it!
I’ve come across some very compelling articles about how to use ChatGPT to streamline the design research process by synthesizing raw interview data to identify themes. Some developers are using it to analyze their code and create documentation. Cool!
This isn’t nearly as creative, but I found it very useful and efficient…
First, Some Background
I love to travel. Whenever I’m heading to a new destination, I spend a bunch of time trying to decide a few things, e.g.
- Where are the best neighborhoods to stay?
- What are the best accommodations in those neighborhoods?
- Which restaurants should I visit?
- What attractions and activities should I prioritize?
This research process typically involves a lot of manual searching and reading (or watching) travel guides and reviews. The problem is everyone who is creating this content is different – their tastes, style, attitudes, etc. are unique and color their experience and preferences. The creators of this content don’t know me or what I prefer, so they can’t tailor their reviews to what I may enjoy. In order to counteract that natural bias, I spend more time searching and reading to develop a wider array of opinions. I look for patterns from several information sources to develop a level of confidence in my decisions. Am I overthinking this? Probably, but I’d rather invest the time up front than end up wasting time and money during the trip having a bad experience.
The ChatGPT Solution
After spending way too mjuch time reading restaurant reviews for an upcoming visit to Nashville, I decided to change my approach. In order to find recommendations tailored to my tastes I fed the following prompt into ChatGPT:
“Here are a few of my favorite restaurants in Indianapolis: Bluebeard, Beholder, Vida, and Tinker Street. Based on that, can you name a few restaurants I’d enjoy in Nashville?”
The results I got back literally made me jump out of my chair and do a victory lap around my dining room table. ChatGPT gave me suggestions for five restaurants in Nashville — three of which I’ve dined with before on previous visits and absolutely loved (in bold below):
Based on your favorite restaurants in Indianapolis, here are a few restaurants in Nashville that you might enjoy:
1. Rolf and Daughters: This award-winning restaurant in Germantown features a creative and seasonal menu of contemporary American cuisine. The industrial-chic atmosphere and open kitchen add to the lively vibe.
2. Henrietta Red: This upscale seafood restaurant in Germantown offers a menu of inventive dishes made with sustainable ingredients. The atmosphere is elegant yet relaxed, with a focus on Southern hospitality.
3. The Catbird Seat: This intimate, chef-driven restaurant in the Gulch offers a unique tasting menu experience with an open kitchen and interactive chefs. Reservations are required, but the experience is worth it.
4. Butcher & Bee: This East Nashville restaurant serves up creative, Mediterranean-inspired dishes made with locally sourced ingredients. The atmosphere is casual and welcoming, with a focus on community and sharing.
5. Lockeland Table: This neighborhood restaurant in East Nashville offers a menu of elevated Southern comfort food, along with a great selection of cocktails and craft beers. The atmosphere is cozy and inviting, with a focus on local ingredients and hospitality.
From there I requested a few more casual dining options and received five more recommendations. In just a few seconds, I was able to use AI to generate ten restaurant recommendations based on my taste rather than the subjective reviews of people who may or may not enjoy similar dining experienced.
I have several more trips planned for the year, and, in addition to restaurants, I plan to use this method to assess potential lodging and activities. If my first experiment holds up, I won’t need to waste hours looking at Yelp, TripAdvisor, or Google reviews for my future travels 😎
