GPT Store Just Launched — What It Means for Consumer and Enterprise GenAI
OpenAI finally launched their GPT Store last week, a few months after the big announcement at OpenAI Dev Day back in November 2023. At Dev Day, I certainly had the impression that this was going to be the end for many in the GenAI ecosystem, but as someone working in this space, my thoughts have most certainly changed in the past two months. Here’s how I’m thinking about GPT Store now.
Would GPT Store disrupt GenAI startups? Probably not.
I am going to make a wild guess that the GPT Store will have almost no impact on most GenAI startups and products in the market that provide any material value-add on top of the GPT APIs. Most of them are (and should be) too complicated to fit into the GPT Store.
What’s GPT Store’s potential for consumer GenAI products?
Here comes the interesting part — It’s definitely too early for conclusions with only three days into the launch, but I also see limited potential for any ground-breaking impact in the consumer GenAI landscape to happen with GPT Store. Here are some of the reasons why:
UI/UX and Functional Limitations: “GPT Girlfriends”
I’ll draw a funny example of the hundreds of “GPT girlfriends” that proliferated the GPT Store to illustrate this point. It’s powerful to have GPT bots inside dating games, where conversations are just one part of the whole experience. However, if you only build a GPT girlfriend bot and make that into a product, the customer experience becomes lackluster. Talking to GPT girlfriends feels like going on endless awkward first dates that are just going nowhere (reminds me a little of my early 20s…). The lack of UI and custom features limits the delivery of any nuanced and personalized customer experiences, which is a quintessential part of most consumer-facing products and apps.
Lack of Product-level Integration: AllTrail GPT
For any chatbot that is related to existing products or services, tying the GPT chatbot to the current user experience and customer problems is very crucial. Here’s a question that every company should ask themselves before GPT Store: Can a chatbot solve your customer problem without access to any of the customer’s information and integration to current workflow?
As a big fan of Alltrails, I was very excited to try out the AllTrails GPT when I saw it placed at the forefront of the GPT store, but a little less excited to see what I actually got. Alltrails GPT seems like a long-winded way to get the information about the trails, which one could’ve easily get in 2 seconds from their already very strong iOS app. It’s not faster for me to get this info from GPT, there’s no additional information given, and I can’t bring the search results back to my app either, so… why GPT?
Because of the lack of product-level integration, I currently don’t see GPT Store giving way to companies like AllTrails to solve any key customer problems. However, it feels to me that this could be a very interactive way to introduce companies and their service to those who haven’t heard of them.
A Niche for Productivity and Functionality-Oriented Tools
Despite all the limitations, GPT Store could carve out a niche and become the battlefield for productivity and functionality-oriented tools. Tools like PDF reader/extractors or research paper summarizer which focuses on narrow and straightforward use cases without the need for deep UI or app-level integration, might be what eventually survive and thrive on the GPT Store.
What about Business and Enterprise GenAI?
For most business/enterprise use cases, the amount of customization and integration needed makes the GPT Store unsuitable. For Retrieval-Augmented Generation (RAG) knowledge bases, companies often need the ability to sync multiple data sources, create different access layers for users, and for users to customize data sources. For customer service chatbots, you need to integrate it into existing UI and draw information from databases. For any sort of process automation use cases, you will need it to be automatically triggered, or fit into the current workflow in whatever way users desire. Most of these sound like something you can’t do on the GPT Store now or even in the near future.
Conclusion
It’s definitely too early for a verdict with only three days into the launch, but as it stands, I think GPT Store is less likely to revolutionize the GenAI landscape or how we think about consumer and enterprise GenAI applications than many have anticipated. With that said, I’m still very curious and excited to see what’s next for the GPT Store from OpenAI!
