ChatGPT is a large language model, an AI that is capable of answering surprisingly complex questions in human language via a chat interface. It is built by OpenAI, a company incorporated in the USA and was released in November 2022, with now more than 100 million users worldwide.
For companies, ChatGPT offers a wide range of use case automation for processes involving human communication and reasoning. Examples include industry-agnostic processes like capital-efficient scaling of customer services, as well as industry-specific processes such as augmentation of high-value technical underwriting in insurance.
In scenarios with sensitive data, it is not recommended to integrate ChatGPT directly. Instead, this booklet showcases ways to integrate open-source alternatives for self-hosted and confidential AI models. Further, ways to anonymize potential inputs for ChatGPT are shown as a way to increase security at scale. Lastly, it shows how building an intellectual training data property can help to avoid lock-in to OpenAI, while still benefiting from the advancements in Artificial Intelligence.
Understand why advancements in textual AI are happening now, and why the processing of human language is something essential for businesses.
Find out more about first use cases of large language models such as ChatGPT, and how they can be used in your business.
General best practices
Dive into open-source alternatives to ChatGPT, how to apply them for your business, and how to understand ChatGPT as one ingredient to a larger puzzle.
Technical best practices
If you want to dive deeper into the technicalities and how to build the training data for your use cases, look into this chapter.
Risks and Obstacles
Understand the risks and obstacles of using ChatGPT and other large language models in your business.
Find material for further research, e.g. in our community chat, YouTube channel or free online courses.
Find a list of terms and definitions that are used throughout this guide.