betterdocs domain was triggered too early. This is usually an indicator for some code in the plugin or theme running too early. Translations should be loaded at the init action or later. Please see Debugging in WordPress for more information. (This message was added in version 6.7.0.) in /home/contentbot/public_html/documentation/wp-includes/functions.php on line 6121In this article we explore how the AI generates the content you receive<\/p>\n\n\n\n
The content generated by ContentBot is synthesized from a large known dataset. The AI models that are used within ContentBot are called Generative Pre-trained Transformers (GPT).<\/p>\n\n\n\n
In order to generate high-quality output, it is necessary to provide these models with high-quality input. Inputs will be transformed into data that the model can understand, allowing it to create original content which is highly relevant to the subject you are generating content for.<\/p>\n\n\n\n
The datasets used to train the AI models should not be thought of as a table of content that is reused but instead can be considered a blueprint for how words and subjects relate to each other. This means our AI models understand subjects it has learnt about.<\/p>\n\n\n\n
At no point does the AI model “scrape the web” or copy other articles – it pulls from its current understanding on that subject and forms its own words, creating unique content based on your inputs.<\/p>\n\n\n\n
We are currently using two AI models: OpenAI and other service providers.<\/p>\n\n\n\n
OpenAI<\/strong> is a powerful GPT-3 model that is capable of producing amazing quality due to the fact that it has one of the largest datasets on the market. We use OpenAI for all short-form content, as well as most of our long-form content.<\/p>\n\n\n\n Other providers are used for some of our long-form content.<\/p>\n\n\n\n Our systems automatically use the best models for the given tool\/task, ensuring that we provide the best quality output for your use case.<\/p>\n\n\n\n Sometimes the AI comes back with weird and wonderful (and sometimes sardonic) results. If you run the tool again it normally comes back with a solid set of copy.<\/p>\n\n\n\n We have also developed quality assurance systems which will attempt to validate the quality of content before outputting it. If our system detects low-quality content, it will automatically regenerate the content.<\/p>\n\n\n\n However, sometimes it just doesn’t know enough on a particular topic to create anything of value. If this happens, get in touch with us (using the chat icon) and let us know. We periodically train the AI to understand new topics and niches.<\/p>\n\n\n\nThe AI created non-sensical content, what now?<\/strong><\/h2>\n\n\n\n