Module: InterRed ContentAgents
SmartAI & Computer Aided Writing
When creating a text, editing options such as "copy & paste", formatting and a spell checker are taken for granted today.
GPT changes the way we deal with text
Systems such as ChatGPT, GPT-3 and Aleph Alpha promise to provide much more advanced support for text creation. And great progress has also been made in the transcription of videos and audio recordings such as podcasts or the tagging or generation of images (DALL-E etc.).
What the new methods based on so-called LLMs ("Large Language Model") have in common is an intrinsic unpredictability of the results. This is problematic mainly because it is also unpredictable whether the result is correct.
GPT stands for "Generative Pretrained Transformer". A GPT like GPT-3 is based on a relatively simple principle: it can generate the most probable "next word" for a given text based on countless learned texts and sentence patterns. Then it again uses the given text and the first generated word to generate the next word. And so on.
In this way, a GPT is able to generate text to a given question that has a high probability of being a suitable answer. Or to a text, to which the sentence "Short this text" is added, to generate a shortened version of this very text.
Why are the results of GPT so unreliable?
A GPT - like everything in the world of computers - is actually deterministic. Therefore, the answer to a given question would always be the same; even verbatim. Why then is GPT-3 not deterministic? As described, a GPT always generates the "next word" with the highest probability. Since there are many words that are very close to this highest probability, this is exploited to introduce "creativity" into the behavior of GPT. This is done by randomly choosing one of the words with the highest probability. At first, this makes little difference, but since this small breeze of randomness is also added to each further generated word, even a sentence with only 10 words and the assumption that there are ten options to choose from for each word already results in "1010", or a billion, variants. In this way, a GPT can generate countless texts on a topic.
Why does GPT not give sources?
If one lets a GPT work on the basis of a given text and little "instruction", the results are usually good. The smaller the basis for text to be generated, the less reliable the result. Particularly problematic is that a GPT does not specify a source for generated text. Why is this?
A GPT has learned all information and its ability to generate text based on an incredibly large amount of texts. Generated text is therefore never based on one source, but on all texts that are relevant in the given context. And relevant in terms of content and text structure.
This can be thought of as similar to us humans. We would also not be able to cite the sources for all our statements. And this is not only because our memory is limited, but also because we base a statement on many sources that we have taken note of in the course of our lives. There identical information strengthens and contradictions weaken each other.
SmartAI - reliable AI that respects boundaries and rules
InterRed ContentAgents are based on state-of-the-art AI technology. However, a unique selling point is the reliability of the methods used. The AI of ContentAgents always respects the set limits and rules. GPT is only used in the form of self-correcting GPT. And generated texts always contain the indication of the underlying sources.
A self-correcting GPT checks compliance with rules, form and content. In response to prompts to describe three important features of self-correcting GPT with one keyword each, a self-correcting GPT would generate exactly three keywords and no more or less, and also exactly one word and not more. Moreover, for each keyword, it would indicate which source the keyword is based on.
We call our AI that adheres to rules and boundaries and documents the underlying sources for generated content "SmartAI."
Efficiency and reliability
With ContentAgents, AI can be used for any form of computer aided writing without uncertainty. Thanks to SmartAI, there is no need for the sometimes time-consuming checking of the AI result, which otherwise often eats up the time gained.
The InterRed ContentAgents support your work with content
As a basis for your work, the ContentAgents automatically analyze all content in the InterRed ContentHub. Any other sources can be added. The analysis is performed completely autonomously on the basis of the predefined sources.
In the process, the InterRed ContentAgents recognize the topics based on the connections in the individual texts. They learn independently, for example, to distinguish topics about a (parking) bank from topics about financial institutions (banks) and to link them with other, suitable content. For example, they would assign articles about money management to the topic "bank" (as the financial institution), but never to the topic "bank" (as a park bank).
InterRed ContentAgents also recognize people and places and provide more information about them. They can also contextually classify and associate synonyms, i.e. associate words like "sun" or "warmth" with "summer".
The fact that they learn independently, constantly readjust themselves and automatically pick up on and incorporate new topics makes them far superior to manual, hierarchical classification systems such as "tagging," ontologies, categorizations or keyword directories. In daily practice, one quickly finds that AI delivers significantly better results than - sometimes at great expense - manually maintained and updated classification systems.
Concrete benefit: Automation
InterRed ContentAgents make AI usable in all areas of everyday editorial work. Below we provide an overview of the available ContentAgents.
Automated print production - SmartPaper
All processes for creating a print product can be automated with SmartPaper technology. From the selection of suitable content, the choice of appropriate images and image sections, to the layout and the shortening of text. In this way, a complete issue can be created automatically within a few minutes. And thanks to SmartAI, which strictly adheres to specifications, the face of the product is preserved and the result does not have to be laboriously checked.
Recommendation of related content - On websites and in everyday work
Whether in everyday editorial work or at the touchpoint for the target group. ContentAgents offer a powerful recommender system that can recommend texts and images that match the content. Suggestions for texts to be linked or suitable images can be integrated into a text simply by dragging and dropping. And at the push of a button, the AI can generate SEO-optimized links based on SEO data that is up-to-date at all times. The optimal teasers can be played out in mobile and web. You set the target. Whether reach, read duration, paywall conversion or product purchase, the ContentAgents support the implementation of the respective strategy.
SmartCollections instead of manual work - use content automatically
Instead of laboriously selecting texts and constantly adapting them to new circumstances over time, SmartCollections make it very easy to define a group of texts with AI assistance. These can then be used as desired. For example, for teasers, a newsletter or part of a print product. As soon as new texts are created that fall within the definition of the SmartCollection, the usage - no matter where - is automatically updated. Because of this feature, SmartCollections are also suitable as a "content radar". You can have the AI automatically inform you about new texts. Via e-mail or message in Microsoft Teams, for example.
FlowEditor - writing in the flow
The integration of ContentAgents into the InterRed FlowEditor makes AI usable without fuss. This means that headings, subheadings, bulleted lists, pre-texts, teaser texts, summaries and social posts (Twitter tweets, etc.) can be generated at the push of a button while writing in the editor. And thanks to SmartAI, always reliably according to your own specifications.