GPTs — Everything you need to know

Are they really the future?

GPTs, as visualized by DALLE3

GPTs are here, GPTs are there, GPTs are everywhere.

Some people call this the opportunity of a lifetime — think of Apple when it first launched its app store. Is the GPT store in the same realms?

For the unheard, here’s a quick recap of OpenAI’s latest update, GPTs.

• Custom GPTs: Create personalized ChatGPT versions for specific purposes.
• Easy Creation: No coding needed; build for personal, company, or public use.
• Example GPTs: Available for ChatGPT Plus and Enterprise users.
• GPT Store: Launching soon for public sharing and monetization.
• Real-World Tasks: Connect GPTs to APIs for real-world integration.
• Enterprise Use: Deploy internal-only GPTs for specific business needs.

I have shared my experience of using and creating GPTs in this LinkedIn post. Here’s a quick summary of it.

1. GPTs are most useful when customized with private data and APIs, especially for business use.
2. Generic GPTs are easy to build but lack value since ChatGPT is widely available for such tasks anyway.
3. Personalizing GPTs for specific tasks enhances productivity, acting like personalized assistants.
4. I created a “PostGPT” to craft social media content in my style. I plan to make more private GPTs for tasks like product management and coding and automate some of my workflows.
5. GPTs that perform real-world tasks or interact with physical devices, like smart home systems, are worth sharing publicly.
6. Digital-only products using GPTs aren’t very effective unless they’re highly personalized or you have a large audience or exclusive data.

Creating a GPT is as simple as conversing with the GPT Builder. You can just paste your prompts and instructions, upload your private data, and then integrate with some external APIs or APIs of your own, and that’s it. You are done.

If you get stuck at any step, you can ask GPT how to proceed. It won’t take more than ten minutes to make a generic digital GPT with access to some out-of-the-box private data files. If you are looking at building a custom backend layer and API suite to get commands for real-world actions from the GPT interaction, that will actually set you apart and provide real value to your users.

These GPTs might be worth a shot to succeed in the GPT store.

As of today, I find GPTs to be a great way to organize our personal workflows rather than a consumer-facing product that you can build for others to use. If you have enough distribution and niche use cases with lots of private data, it makes sense.

But otherwise, GPTs are like your private army of assistants that can supercharge your productivity at a personal level. Using someone else’s GPT feels counterintuitive when you can simply build your own and customize to your unique style and requirements.