Systems such as ChatGPT, GPT-3, GPT-4 and Aleph Alpha promise significantly 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 indexing or generation of images (DALL-E etc.).

What the methods of the new generative AIs, based on so-called LLMs ("Large Language Models"), have in common is the inherent unpredictability of the results. This is particularly problematic because it is also impossible to predict whether the result will be correct.

GPT stands for "Generative Pretrained Transformer". A GPT such as GPT-4 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. It then uses the given text and the first word generated to generate the next word. And so on.

In this way, a GPT is able to generate a text for a given question that has a high probability of being a suitable answer. Or to a text to which the sentence "Abbreviate this text" is added, to generate an abbreviated version of the same text.

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. So why is GPT-4 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 used to introduce "creativity" into the behavior of GPT. To do this, one of the words with the highest probability is chosen at random. Initially, this makes little difference, but since this little touch of randomness is also added with each additional word generated, a sentence with only 10 words and the assumption that there are ten options to choose from for each word already results in "¹⁰¹⁰", i.e. a billion, variants. In this way, a GPT can generate countless texts on one topic.

If you let a GPT work on the basis of a given text and few "instructions", the results are usually good. The smaller the basis for the text to be generated, the less reliable the result. It is particularly problematic that a GPT does not specify a source for the generated text. Why is this the case?

A GPT has learned all the information and its ability to generate text based on an incredibly large amount of text. 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.

You can imagine this in a similar way 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. Identical information reinforces each other and contradictions weaken each other.

×

InterRed ContentAgents are based on state-of-the-art AI technology. However, a unique selling point is the reliability of the processes used. The AI of the ContentAgents always observes the set limits and rules. GPT is only used in the form of self-correcting GPT. And generated texts always contain information about the underlying sources.

A self-correcting GPT checks compliance with rules, form and content. When asked to describe three important properties 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 several. It would also indicate for each keyword which source the keyword is based on.

We call our AI, which adheres to rules and limits and documents the underlying sources for generated content, "SmartAI".

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 AI results, which otherwise often eats up the time saved.

The ContentAgents automatically analyze all content from the InterRed ContentHub as a basis for your work. Any other sources can be added. The analysis is carried out completely autonomously based on the predefined sources.

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 the money economy to the topic "bank" (as the financial institution), but never to the topic "bank" (as a park bank).

The InterRed ContentAgents also recognize people and places and provide further information about them. They can also contextually classify and associate synonyms, i.e. associate words such as "sun" or "warmth" with "summer".

Because they learn independently, constantly readjust themselves and automatically pick up and incorporate new topics, they are clearly superior to manual, hierarchical classification systems such as tagging, ontologies, categorizations or keyword directories. In day-to-day practice, it quickly becomes clear that AI delivers significantly better results than manually maintained and updated classification systems, which can involve a great deal of effort.

The InterRed ContentAgents make AI usable in all areas of everyday editorial work.

All processes for creating a print product can be automated with SmartPaper technology. From the selection of suitable content, the choice of suitable images and image sections to the layout and shortening of text. A complete edition can be created automatically in just a few minutes. And thanks to the SmartAI, which adheres strictly to specifications, the face of the product is retained and the result does not have to be laboriously checked. Among other things, InterRed SmartPaper uses the artificial intelligence of InterRed ContentAgents and thus the InterRed SmartAI technologies for automation.

Whether in everyday editorial work or at the touchpoint for the target group. The ContentAgents offer a powerful recommender system that can recommend related texts and images. Suggestions for texts to be linked or suitable images can be integrated into a text simply by dragging and dropping. And at the touch of a button, the AI can generate SEO-optimized links based on up-to-date SEO data at any time. The optimal teasers can be displayed on mobile and web. You set the target. Whether reach, reading time, paywall conversion or product purchase, the ContentAgents support the implementation of the respective strategy.

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 the help of AI. These can then be used as required. 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. This feature also makes SmartCollections suitable as a "content radar". You can have the AI automatically inform you about new texts. By e-mail or message in Microsoft Teams, for example.

The integration of ContentAgents into the InterRed FlowEditor makes AI easy to use. Headings, subheadings, bulleted lists, pre-texts, teaser texts, summaries and social posts (Twitter tweets etc.) can be created at the touch of a button while writing in the editor. And, thanks to SmartAI, always reliably according to your own specifications.