<aside> 📌
proofread by ai; should not have mistakes
</aside>
Today, I was in the mood for some Chinese food after work and was looking at a couple of nearby restaurants to try out. But when hunger strikes, you need to make decisions fast – ideally, have food in front of you within minutes. Still, I wanted to try authentic Sichuan cuisine and experience bold flavors and spiciness in every bite.
Normally, I start by searching for restaurants, checking their menus, and reading reviews. If someone hasn’t personally recommended a restaurant to me, I do my own “due diligence” before picking one. If I’m grabbing a quick meal alone, I definitely want to try their best dishes. If I’m out with friends, we usually order different items to share so we can try everything.
But alas, I wanted to check out a couple of dishes for myself and found myself combing through countless reviews. After what felt like “one billion years,” I finally decided on a restaurant and what I would try there. I headed out to enjoy my cheat day. The staff was amazing, the vibe enjoyable, and the environment welcoming, with lots of families enjoying an early dinner.
While waiting for my food, I started wondering - why do users have to spend so much time answering such basic questions?
While I eventually pieced together answers from different reviews, I realized I had wasted a lot of time - and my hunger prompted me to imagine something better.
So while waiting and watching happy families enjoy their meals, I started thinking about an update to Google Places reviews to enhance user experience using Gemini.
In my months of using the Gemini API and being consistently impressed by its context length and capabilities, I’m confident that something like this is absolutely possible.
Alongside the standard list of reviews, users could query a chatbot that uses the past few months of reviews as context.
