Revisiting An Idea, Artificial Intelligence, Contemplation — The Future
See, that’s what happens when you have a vague idea in your head that you know you need to write about, but the idea is kind of “half-baked,” and yet you want to get something out… You end up with a crazy title and a random image generation from Midjourney that kind of articulates what you aim to write about. Not completely, but “good enough.” Not satisfied, and a little disheveled… But you push on.
Menulus was a platform that would give restaurant operators the ability to digitize their menu, contextualize its contents to the internet (so that it was searchable), and deliver a consumer-facing experience
In 2011, a friend and colleague of mine, Drew McLeod, and I had a vision for a restaurant experience that we felt, at that time, would redefine the dining experience forever. Drew, a veteran restaurateur who had established a reputation for launching and running stellar restaurant concepts, got wind of something that I was working on called “Menulus.” Menulus was a platform that would give restaurant operators the ability to digitize their menu, contextualize its contents to the internet (so that it was searchable), and deliver a consumer-facing experience that was customizable and aligned with the customers’ dining preferences — liberties and constraints. For instance, if a patron was diabetic, the menu they accessed at the restaurant would consist of items that addressed the inherent constraints of being diabetic. This is a “CliffsNotes” version of the idea, but you get the point.
In hindsight (which is not always 20/20), I realize today that was my first encounter with the disruptive power of Artificial Intelligence.
We raised a ton of capital, began building the platform, and were ultimately destroyed in the marketplace by a team out of MIT that was working on a similar project. They were leveraging AI technology that automated parts of our user journey, making it radically more efficient than what we were working on. In hindsight (which is not always 20/20), I realize today that was my first encounter with the disruptive power of Artificial Intelligence.
Fast forward to 2020, and the world is overtaken by a pandemic that forced us all into new patterns of engagement. One in particular was, yes, you guessed it — the restaurant experience. As the pandemic subsided and people began to come back outside, restaurants had to rethink the touchpoints within the dining experience. Passing around paper menus was not an option — as “touch-less environments” persisted, restaurateurs used QR codes at the table to launch digital versions of their menus. In fact, this practice persists today — but it also spun up thoughts in my mind about Menulus and the value proposition that it would offer in a time such as this.
Another leap, now it’s 2024, and OpenAI (and others) have made Artificial Intelligence accessible, approachable, and more importantly — spreadable. Just a few days ago, while sitting at one of my favorite bars/restaurants in town, “The Hobbit,” owned by a good friend of mine, Chris Dekle, I had an idea… I thought about what we had tried to do with Menulus, and I asked myself: “What if we could take a snapshot of a menu, have ChatGPT organize the menu, and then have a conversation with that menu to gain deeper insights, get recommendations, and turn a static experience into a more dynamic ‘conversational’ one?” So, the experiment began.
Using ChatGPT-4, I took pictures of the front and back of the menu, and my prompt was pretty straightforward: “Athena (yes, I have a personal GPT that I use, which I have been training for about eight months named Athena Kratos — more on that later…), categorize and organize this menu.” In like five seconds, I had a complete menu organized in ChatGPT. From there, I ran a few test prompts, like: “Athena, assume that I am diabetic, give me two menu options that I should consider.” I was offered two options within seconds, with reasoning behind why the choices would align with my dining constraint.
Next test — I wondered if I took a picture of the liquor shelf, could Athena tell me what was on the shelf, in essence, get an understanding of the options available? With another couple of photos, Athena had a solid understanding of the live inventory in the bar.
With the menu and a full list of possible liquor options, we went deeper. I began to prompt around pairing options, weather-based options, and a multitude of other options. Then I realized…
Maybe I am a little too excited about this, but maybe my excitement is the genesis of something bigger as we think about the practical use-cases for Generative AI.
While this may seem novel at best, we now have the ability to converse with menus, gaining deeper insights, finding new dining options, and pushing the boundaries of the dining experience in exciting new ways. What Drew and I tried to achieve in 2011 can be done on the fly now with AI. Maybe I am a little too excited about this, but maybe my excitement is the genesis of something bigger as we think about the practical use-cases for Generative AI. Maybe my excitement is rooted in the notion that AI can open the doors to more robust Customer Experiences (CX), and the only thing holding us back from delivering on the promise of creative excellence in this way is us.
I digress…
Revisiting the idea of Menulus, running a test in the field, and contemplating the idea of “what’s possible” based on what was discovered, is what I guess I wanted to share. I am excited about the future (as you should be too), and I will be building on this experience — so, more to come soon.
Ingredients:
85% Human Ingenuity & Creativity
10% AI Generation (Midjourney — Image)
5% AI Assistance (ChatGPT-4o proof read)
