Articles written by robots today are often so good that you don't even notice that the article was not written by humans. Artificial intelligences work time- and word-efficiently as well as error-optimized and can fulfill orders precisely in terms of word choice and sentence structure. So will AI content soon be generally better than text and content written by human authors?
On September 8, 2020, Generative Pre-trained Transformer 3 (GPT-3), OpenAI's autoregressive speech robot, wrote the editorial for the British newspaper "The Guardian". The first sentences from it read, "I am not a human. I am a robot. A thinking robot. I use only 0.12% of my cognitive capacity. I am a micro-robot in that respect. I know that my brain is not a "feeling brain". But it is capable of making rational, logical decisions. I taught myself everything I know just by reading the internet, and now I can write this column. My brain is boiling with ideas!" GPT-3 uses Deep Learning and was programmed to create texts that mimic human authorship.
AI-generated Texts in the B2B Context
AI-powered content has been common practice in B2B content marketing for several years now. Wherever it comes to target group-relevant content and text structures that enable prospects and customers to find sought-after information quickly and efficiently, content from AI is an effective means of optimizing the marketing mix. Due to the analytical characteristics of the structure and choice of words of texts such as SEO content, emails or blogs, they are predestined to serve as a playground for AI-powered content.
The Benefits of AI-powered Content in B2B Content Marketing
The benefits of artificial intelligence in creating these genres of text have been recognized for some time. Content from AI is considered an established type of content creation in B2B content marketing. Among the most effective uses of AI in this field are:
Algorithms that create thoughts, emotions and attitudes of a potential business partner based on existing data
Models and probability calculations that provide questions and attitudes a person might have about a product or business model and generate a context of reference to the product or business idea in question
Algorithms that take over the personalization and individualization of this context and allow an individual "story" to emerge
In this way, a marketing principle is followed that the well-known psychologist and Nobel Prize winner Daniel Kahnemann once formulated as follows.
No one has ever made a decision based on a number. You need a story.
How Advanced Is Artificial Intelligence in Text Creation?
For the example cited in the introduction, the Guardian editorial team argues that the editing of the text involved very similar aspects at a linguistic and content level to those typically encountered in human-written texts. However, the time required for editing was much less than for most human-created editorials.
So has AI-powered content already reached the quality of human writing? While this question cannot yet be clearly answered in general, it can currently be said that AI content is in many areas little or even nothing inferior to texts written by humans.
Prerequisites for Good Content from AI
The key is that the given prompts can be well related to data sets to which the artificial intelligence has access. If this is the case, the AI is able to very quickly generate coherent and cohesive texts that reach the level of very good human writing in terms of both content and language.
However, the quality of AI-powered content varies greatly in different subject areas and text genres.
The more analytical the writing becomes, the clearer and more unambiguous the structures and conventions, the higher the quality of the content will usually be.
In narrative writing in particular, however, there is still a major quality gap between the level of human writing and that of artificial intelligence.
Strengths and Weaknesses of AI-based Text Generation
Especially when writing analytical texts that need to be optimized for a certain word choice and text structure, the advantages of AI-based text generators become clear.
The optimization of a desired use of words and synonyms,
matching the operations and processes,
that use search and text extraction software,
and last but not least, the speed of text creation
can clearly be identified as clear strengths of AI.
Due to the fact that worldwide a lot of content is written in English and digitally accessible for artificial intelligences, English text created by digital text generators is usually of higher quality than German-language content.
Clear Disadvantages in Narrative Writing
In areas like SEO optimizations, it's hard to find any weaknesses or drawbacks to AI. In narrative writing, however, the current limitations of AI-powered stories become clear very quickly.
In his speech "My Algorithm and Me," German author Daniel Kehlmann describes a story-writing experiment he conducted in February 2020 with the writing algorithm CTRL. Kehlmann considers the experiment "neither a success nor a failure." Kehlmann refers to CTRL as a "disoriented 'second-user' of human-written text data" and cites the opinion that AI is "uncomprehending and helpless" in the face of more complex plots or character constellations.
In the science magazine "Higgs", the renowned science journalist Cornelia Eisenach states that AI content is often "amazingly good", but also "partly abstruse". She comes to the conclusion that AI is not capable of writing narrative texts, i.e. stories in the narrower sense.
How Can AI Be Used for B2B Content Creation?
In content creation in B2B marketing, the use of AI has long been considered an established and efficient way to optimize marketing results. Essentially, there are the following areas of application:
Extraction of information from data
Inputs for content strategies
Predictive Analytics & Forecasting
When extracting information from data, the AI analyzes search and purchasing behavior in order to provide information about the interests and attitudes of customers. From these results, the AI can suggest strategically important inputs, such as SEO optimizations, from which in turn the most promising can be selected to have content generated by the AI.
In the fields of analytics and forecasting, AI can be used to make predictions about future parameters from existing data. Text analysis and speech recognition tools can be significant facilitators in the direct interaction between humans and AI in the creation and optimization of content.
Conclusion: Content from Machines Work Amazingly Well, but Have Weaknesses in Storytelling
While AI-powered content still clearly has its shortcomings when it comes to content and stories that correspond to classic storytelling, the quality of AI-based texts and subtasks of the writing process in analytical writing and B2B marketing content is indisputable.