- Nexwave
- Posts
- We Built Our Own GPT Through OpenAI, and Here’s How It Went 😎
We Built Our Own GPT Through OpenAI, and Here’s How It Went 😎
Exploring OpenAI's latest product launch, GPTs!
Kendall Rankin & Chad Payne
November 14, 2023
Happy Tuesday, everyone!
While some of us might still be riding the Halloween sugar high, brace yourselves for the Mariah Carey anthems, Christmas hype, and yes, the bone-chilling cold (seriously, its brick out there!). But have no fear, because we've got some exciting AI news to warm you up (pun intended).
OpenAI just unwrapped an early Christmas present – GPTs. It's like DIY ChatGPT, and you don't need to know any code to do it! To show you how it works, we’re putting GPTs to the ultimate holiday test: nailing that signature holiday recipe the family won't stop raving about.
Let's dive into the tech magic and sprinkle a bit of AI into your holiday prep! 🎉🤖
Let’s start with the basics. What are Large Language Models (LLMs)?
If you’ve tried any of the popular AI chatbots out right now, like ChatGPT, or Bard, or Claude, to name a few, you’ve probably asked yourself “how does this work?” It’s a fair question, and one we’ll answer for you. These popular programs are large language models (LLMs), AI systems that are trained on massive amounts of data ranging from internal documents, to public forums like Reddit and Wikipedia.
Using specific neural network architecture, like transformers, LLMs can process text data and identify patterns, allowing them to make predictions about what words should come next. Each transformer has a self-attention mechanism that allows for that sophisticated understanding of the relationship between words in a sentence. Because it’s more relational, every output from an LLM is different; but that also means there’s a lot of room for error, and oftentimes super cliche or generic text. So human intervention is still required to train and supervise these programs.
But what if these LLMs were built around exactly what you wanted. What if they studied your interests and data and were custom made to answer your questions, whatever they may be? Well, that’s not a “what if” question anymore…
OpenAI just changed the game, and here’s how.
We’ve set the stage by giving you the rundown on what LLMs are and how they work, and now it’s time to talk about the announcement that’s taken the tech world by storm.
Last week, OpenAI launched GPTs – a product, powered by AI, that gives anyone, anywhere the ability to create their own, highly customizable versions of ChatGPT that help you achieve a specific purpose. And yes, we really do mean anyone, including people like us and you! (If you don’t believe us, keep scrolling and we’ll show you just how easy it is 🤝🏽) Want to know the best part about it? Using their product, you don’t have to know how to code to create your own GPTs.
Let’s just say the possibilities are endless. OpenAI is leading the way in creating accessible and responsible AI, and the launch of customizable GPTs that don’t require any coding is a huge step in the right direction.
So who should be creating GPTs and what’s their value prop?
No matter who you are or what you do, there’s a way for you to create your own GPTs tailored exactly to your specific needs. Are you a business owner with limited resources who wants to integrate a customer chatbot that manages itself into your website? Or an executive looking to post thought leadership on LinkedIn who needs a creative writing coach to take their work to the next level? Or the one person in your family who doesn't know how to play spades and needs a game coach so your family can stop clowning you? Well, you can build a GPT to help you.
But wait – there’s more. 👀 OpenAI's GPTs offer developers the opportunity to create their own conversational AI systems using OpenAI's models and make them available for users through the GPT Store, a marketplace hosted by OpenAI. During their Developers Day conference, OpenAI CEO, Sam Altman, announced that developers will soon have the ability to generate revenue from their GPTs, with earnings tied to the number of people who use their creations.
Why are they the next best thing? Because they are pre-trained, generative, and unsupervised. It’s like having a personal fairy on your shoulder that can coach you through (almost) anything. And the fact that there’s an opportunity for people like us to create GPTs and make money from them is pretty exciting.
We promised you that we’d show you just how easy it is to make your own GPT. Let’s try it out!
Now if you’re based in the United States, you know that Thanksgiving is coming up next Thursday – a time for families and loved ones to get together and enjoy a fire meal. One of the dishes we all know and love that’s most popular (but most controversial if you get it wrong), is none other than mac & cheese. So what better way to test out building GPTs than to see if the mac & cheese recipe from our sous chef is valid or not?!
When you first get to the Editor, you have two panels side-by-side. The one on the left is the GPT Builder where you can tell it what you want to make, and the one on the right is the GPT itself where you can ask questions and gain insight/advice.
We decided to ask the GPT, affectionately named “Cheesy Chef” by the GPT Builder, what types of cheeses work best with sharp cheddar and what seasonings to use. The results? A hit!
It also gave a little extra rizz by suggesting toppings and sharing more tips. Okay, Cheesy Chef! 👏
We followed up asking for a recipe that incorporates all of the ingredients it suggested.
Is this recipe as good as your grandma’s? I guess we’ll have to actually try it to find out! But from first glance, it looks like a pretty damn good recipe.
Back over on the other panel, the GPT Builder suggested a name for it, and even continued to ask questions around what type of tone and personality we’d want it to have. We’re loving the customization.
We’re unashamed to say that after experimenting with building our first one, it’s become quite addicting to see everything else these GPTs can do. This was a simple example designed to show you at a high level how it works and just how easy it is. We definitely encourage you all to give it a try if you have a Plus or Enterprise OpenAI account. Who knows? Take it seriously and your GPT could end up as a hit on the GPT store.
So what does this mean for us?
Our biggest takeaway is that if you are a human who breathes (all of us lol), you likely have a reason to begin exploring how GPTs can make your life easier – whether for yourself or for your company. The era of personalized, trainable AI is here, and the possibilities are endless. Why can’t we be the ones the create them? Let’s get it, family!
Company Spotlight: Character.ai 🦸
On the topic of generating conversations with AI bots, Character.ai is an a16z-backed chatbot startup that announced their closing of a $150M Series A round earlier this year at a $1 billion valuation. Wild, right?!
Plainly put, it’s an AI app maker founded by ex-Google employees that lets users design their own AI characters and have conversations with them. Right now, they’re seeing about 4.2 million monthly active users in the U.S., compared to nearly 6 million monthly U.S. active users for ChatGPT’s mobile apps. Seems like they’re getting a good amount of traction. Test it out and let us know what you think!
Jobs in Web3: Bag Talk 💰
Want to jump on that AI wave? Check out some roles below. And don’t worry, if you’re looking for that internship, we got you too! Also, don’t be a stranger - let us know what types of roles / companies you’re looking for - let’s secure the bag y’all!
What’s Happening in the News 🗞️
Here’s what caught our attention over the past two weeks –
As always, thanks for reading! Hope you learned something new today, because we sure did. 😄
– Kendall & Chad