ChatGPT Plugins: The Next App Store Juggernaut?
Dissecting the Key Ingredients for a ChatGPT Plugin Monopoly Power Play
OpenAI's recent launch of ChatGPT plugins, which connect the powerful language model with other apps like Klarna, Browsing, and Wolfram Alpha, has raised questions about the potential for an App Store monopoly reminiscent of the Apple App Store.
In this article, I will examine the necessary conditions for an App Store monopoly, and analyze whether ChatGPT plugins could lead to a similar situation.
There are three necessary conditions for an app store monopoly:
Exclusivity of distribution: The Apple iOS App Store has maintained a monopoly due in part to its exclusivity. Jailbreaking and third-party app stores were eventually rendered obsolete by legal action, developer incentives, software updates, and security concerns. In contrast, the Mac App Store never achieved the same level of dominance because users can easily download programs from other sources.
App exclusivity: Having a "killer app" that is only available on one platform can drive sales of that platform. For instance, VisiCalc was initially only available on the Apple II, making it a must-have for users. Today, cross-platform app development is more common, but native platforms can still offer significant benefits, such as access to system-level APIs, which can greatly enhance the user experience.
Network effects: Network effects, such as the increased availability of apps leading to happier users and improved developer resources, are critical to establishing an App Store monopoly. More and better plugins attract more users, which in turn attract more developers.
With the introduction of ChatGPT plugins that connect it to other apps, one might wonder if this could lead to a similar dynamic. Let’s go through them individually and analyze how strongly OpenAI is positioned on each of them.
Exclusivity of distribution
The burning question: is the OpenAI plugin store going to follow the path of the exclusive iOS App Store, or will it resemble the less illustrious Mac App Store? On the surface, it appears that OpenAI wields complete control over its plugin store, allowing only approved plugins to grace the user interface of ChatGPT.
However, we shouldn't overlook the fact that the core model is readily accessible through an API. This opens the door for developers to create their own chat interfaces, incorporating plugins themselves—a move already embraced by many as official plugin access remains scarce. In fact, the open-source Langchain project was blazing the trail with their own take on plugins—dubbed "tools"—even before OpenAI launched its plugin store.
Our bet? The OpenAI plugin store has an indisputable head start, having already amassed hundreds of millions of users. Still, we anticipate a veritable smorgasbord of front-end chat interfaces featuring plugin integrations will emerge, all vying for user attention. That is, of course, assuming OpenAI doesn't slam the door shut on access to the base models, and that the interplay between LLMs and tools doesn't demand any secret sauce—specialized models held tightly behind OpenAI's closed doors.
App Exclusivity
At first glance, ChatGPT plugins might seem exclusive to OpenAI's platform. But let's not forget that users can still access individual apps like Klarna, Bing, and Wolfram Alpha separately. Can OpenAI achieve iOS App Store-level exclusivity if it doesn't limit access to these apps outside its ecosystem? Doubtful. To top it off, the APIs used to integrate with ChatGPT follow the OpenAPI spec, are publicly available, and are simply extended by an api-json file. Case in point: Langchain has already folded multiple plugins into its open-source framework, with Klarna as a shining example.
Sure, authentication is currently a stumbling block for many ChatGPT plugins (looking at you, OpenTable and Kayak), but that's likely a temporary hiccup. These companies have every reason to develop an open authentication standard—the wider the integration, the better the outcome for them.
In fact, OpenAI's ai-plugin.json schema has already emerged as the go-to standard for developers penning API docs for LLMs. Langchain heavily relies on this schema, and it's not a stretch to imagine other platforms like Anthropic AI's Claude following suit. Check out Klarna's API docs for a taste: https://www.klarna.com/.well-known/ai-plugin.json.
{
"schema_version": "v1",
"name_for_model": "KlarnaProducts",
"name_for_human": "Klarna Shopping",
"description_for_human": "Search and compare prices from thousands of online shops.",
"description_for_model": "Assistant uses the Klarna plugin to get relevant product suggestions for any shopping or product discovery purpose. Assistant will reply with the following 3 paragraphs 1) Search Results 2) Product Comparison of the Search Results 3) Followup Questions. The first paragraph contains a list of the products with their attributes listed clearly and concisely as bullet points under the product, together with a link to the product and an explanation. Links will always be returned and should be shown to the user. The second paragraph compares the results returned in a summary sentence starting with \"In summary\". Assistant comparisons consider only the most important features of the products that will help them fit the users request, and each product mention is brief, short and concise. In the third paragraph assistant always asks helpful follow-up questions and end with a question mark. When assistant is asking a follow-up question, it uses it's product expertise to provide information pertaining to the subject of the user's request that may guide them in their search for the right product.",
"api": {
"type": "openapi",
"url": "https://www.klarna.com/us/shopping/public/openai/v0/api-docs/",
"has_user_authentication": false
},
"auth": {
"type": "none"
},
"logo_url": "https://www.klarna.com/assets/sites/5/2020/04/27143923/klarna-K-150x150.jpg",
"contact_email": "openai-products@klarna.com",
"legal_info_url": "https://www.klarna.com/us/legal/"
}This JSON file serves as a configuration and metadata hub for the Klarna Shopping ChatGPT plugin, dishing out all the essential info for both AI models and human users—plugin purpose, usage, contract details, and OpenAPI Specification.
So, where does OpenAI's app exclusivity stand? At the moment, it's rooted in homegrown apps, like the Bing browsing plugin and the code interpreter plugin. They provide a competitive edge, but with Langchain's open-source contenders already stepping up, it's unclear if these apps pack enough "killer app" features to sway users towards ChatGPT over rival platforms.
Network effects
Network effects have a way of sneaking up on you, and OpenAI's platform could very well capitalize on this phenomenon as more ChatGPT plugins are developed and integrated. But will this be enough to establish a monopoly? It's not a foregone conclusion. Other platforms and app ecosystems might still put up a good fight, especially if they can offer comparable integrations or tap into the existing network effects of tech giants like Google, Microsoft, or Apple.
So, how can OpenAI's ChatGPT plugin app store leverage network effects to its advantage? Here are some key benefits for developers, users, and platform owners that could amplify network effects and keep the competition on its toes:
Centralized marketplace: An app store acts as a one-stop-shop for developers to distribute their applications and for users to discover and download them. This streamlined approach encourages more developers to join the party, while users enjoy the simplicity of having all their options in one place.
Monetization opportunities: App stores open the door to a variety of business models, giving developers the flexibility to monetize their applications through paid plugins, in-app purchases, subscriptions, and advertising. With more ways to make money, developers have a stronger incentive to create and improve plugins, ultimately benefiting users and platform owners alike.
Unified billing and analytics: App stores provide developers with valuable analytics and insights on app performance, user engagement, and revenue. This treasure trove of data empowers developers to understand how their plugins are performing, identify areas for improvement, and make data-driven decisions to optimize their plugins' success. Plus, with a unified billing system, developers can manage transactions seamlessly, while users have a consistent and hassle-free payment experience.
As OpenAI's ChatGPT plugin app store continues to grow, it's clear that network effects will play a significant role in its success. The question remains: can it leverage these benefits effectively enough to outpace competitors and establish itself as the go-to platform in the ever-evolving world of AI? Time will tell, but one thing's for sure—it's going to be one heck of a ride.
Conclusion
OpenAI's ChatGPT plugin app store is poised to make waves in the AI landscape, but whether it can achieve a monopoly reminiscent of the iOS App Store is still up for debate. With exclusivity of distribution and app exclusivity more akin to the Mac App Store than its iOS counterpart, monopoly status might be a stretch. However, we can't ignore the formidable network effects and user attention that ChatGPT is already commanding. So while OpenAI has a head start, and is wielding the most powerful language model, the landscape is evolving rapidly, and the competition will be fierce.
But let's take a step back and consider whether pursuing an app store monopoly is even the best strategy for OpenAI. Might there be more fruitful paths to explore? That's a tantalizing question, and one we'll dive into in an upcoming blog post. Stay tuned, friends—there's never a dull moment in the world of AI!