The next app store... but for AI.
If you're reading this blog post and you're over 25, you may remember when the "App Store" was introduced by Apple in 2008. The initial release had about 500 registered mobile iOS apps, and today it has over 1.8 million. The role of the App Store has created jobs in the "app economy" and Apple claims the app economy to have paid over $155 billion to developers.
It's easy to say that the invention of the app store has been incredibly impactful in our modern world. So what does this have to do with AI?
At this point, ChatGPT is old news. The public AI model, released by OpenAI, has revolutionized the way people and businesses are using artificial intelligence. They've released GPT-3, 3.5 turbo, and now GPT-4. But last week, on March 23rd, the company announced Plugins for ChatbotGPT!
ChatGPT has introduced support for plugins which are specially designed tools for language models that prioritize safety as a fundamental principle. These plugins aid ChatGPT in accessing the latest information, performing computations, and utilizing external services.
Since ChatGPT's launch, users have been expressing a desire for plugins, and many developers have been experimenting with similar concepts, as they offer a wide range of potential use cases. The introduction of plugins will initially be limited to a small group of users, and as more information is obtained through an alpha period, access will be scaled up for plugin developers, ChatGPT users, and eventually API users who want to integrate plugins into their products. A community of enthusiasts is being established to shape the future of the human-AI interaction paradigm, and ChatGPT is excited to be a part of this process.
Plugin developers who have been invited off the waitlist can use ChatGPT's documentation to create plugins for the platform, which will then be listed in the prompt displayed to the language model. Documentation will also be provided to guide the model on how to use each plugin. Initial plugins have been developed by Expedia, FiscalNote, Instacart, KAYAK, Klarna, Milo,OpenTable, Shopify, Slack, Speak, Wolfram, and Zapier.
In addition, ChatGPT is hosting two plugins - a web browser and a code interpreter - on its platform. The code for a knowledge base retrieval plugin has also been made available for open-source use, allowing any developer to host their own version to supplement ChatGPT with relevant information.
Although language models have proven useful for various tasks, they are still constrained in their capabilities. The only information they can learn from is their training data, which can be outdated and not tailored to specific applications. Moreover, language models can only generate text and require additional processes to execute instructions.
Plugins can be thought of as "eyes and ears" for language models, providing access to information that is too recent, personalized, or specific to be included in the training data. Plugins can also enable language models to perform safe and controlled actions based on a user's request, thus enhancing the system's functionality.It is expected that open standards will emerge to standardize the way applications expose an AI-facing interface.
As an early attempt to establish such a standard, ChatGPT is developing a protocol and seeking input from developers interested in collaborating.ChatGPT is beginning to gradually activate existing plugins from its early partners for ChatGPT Plus subscribers, while also providing developers with the ability to create their own plugins for the platform. As safety systems are improved and insights are gained from deployment, ChatGPT will continue to refine the protocol and eventually make it possible for developers to integrate plugins into their own applications using OpenAI models beyond ChatGPT.
Whats this all mean!?!?!
The introduction of plugins to ChatGPT could potentially revolutionize the app store model by providing a new way for developers to extend the capabilities of language models. Similar to how apps can enhance the functionality of smartphones, plugins can enhance the capabilities of language models, opening up a vast range of potential use cases.
If adopted on a large scale, this could be the next big thing in AI-powered technology, as it would offer a new level of customization and functionality to language models.
Here at Function Factories, we provide services for creating plug-ins for your application within ChatGPT. Along with this, we can host and maintain the infrastructure for you!
It's easy to say that the invention of the app store has been incredibly impactful in our modern world. So what does this have to do with AI?
At this point, ChatGPT is old news. The public AI model, released by OpenAI, has revolutionized the way people and businesses are using artificial intelligence. They've released GPT-3, 3.5 turbo, and now GPT-4. But last week, on March 23rd, the company announced Plugins for ChatbotGPT!
ChatGPT has introduced support for plugins which are specially designed tools for language models that prioritize safety as a fundamental principle. These plugins aid ChatGPT in accessing the latest information, performing computations, and utilizing external services.
Since ChatGPT's launch, users have been expressing a desire for plugins, and many developers have been experimenting with similar concepts, as they offer a wide range of potential use cases. The introduction of plugins will initially be limited to a small group of users, and as more information is obtained through an alpha period, access will be scaled up for plugin developers, ChatGPT users, and eventually API users who want to integrate plugins into their products. A community of enthusiasts is being established to shape the future of the human-AI interaction paradigm, and ChatGPT is excited to be a part of this process.
Plugin developers who have been invited off the waitlist can use ChatGPT's documentation to create plugins for the platform, which will then be listed in the prompt displayed to the language model. Documentation will also be provided to guide the model on how to use each plugin. Initial plugins have been developed by Expedia, FiscalNote, Instacart, KAYAK, Klarna, Milo,OpenTable, Shopify, Slack, Speak, Wolfram, and Zapier.
In addition, ChatGPT is hosting two plugins - a web browser and a code interpreter - on its platform. The code for a knowledge base retrieval plugin has also been made available for open-source use, allowing any developer to host their own version to supplement ChatGPT with relevant information.
Although language models have proven useful for various tasks, they are still constrained in their capabilities. The only information they can learn from is their training data, which can be outdated and not tailored to specific applications. Moreover, language models can only generate text and require additional processes to execute instructions.
Plugins can be thought of as "eyes and ears" for language models, providing access to information that is too recent, personalized, or specific to be included in the training data. Plugins can also enable language models to perform safe and controlled actions based on a user's request, thus enhancing the system's functionality.It is expected that open standards will emerge to standardize the way applications expose an AI-facing interface.
As an early attempt to establish such a standard, ChatGPT is developing a protocol and seeking input from developers interested in collaborating.ChatGPT is beginning to gradually activate existing plugins from its early partners for ChatGPT Plus subscribers, while also providing developers with the ability to create their own plugins for the platform. As safety systems are improved and insights are gained from deployment, ChatGPT will continue to refine the protocol and eventually make it possible for developers to integrate plugins into their own applications using OpenAI models beyond ChatGPT.
Whats this all mean!?!?!
The introduction of plugins to ChatGPT could potentially revolutionize the app store model by providing a new way for developers to extend the capabilities of language models. Similar to how apps can enhance the functionality of smartphones, plugins can enhance the capabilities of language models, opening up a vast range of potential use cases.
If adopted on a large scale, this could be the next big thing in AI-powered technology, as it would offer a new level of customization and functionality to language models.
Here at Function Factories, we provide services for creating plug-ins for your application within ChatGPT. Along with this, we can host and maintain the infrastructure for you!