Last week I had an interesting call with a company that has been asked by IEEE to generate articles based on the material I have produced in the area of Digital Transformation. They have a team that will write the articles and they got in touch with me.
The interesting part is that they did not read my writings in the area. They have an AI based software that looked at my extensive production, in this blog, in ebooks and webinars and extracted what it felt are the key points in my publications (and talks). That resulted in a suggestion of articles to be produced (3) the ideal length for each of them (to the point of indicating the exact number of words (!) and the specific points to be addressed in each articles, plus three suggested titles for each that, according to the software, would generate most hit by search engines.
The purpose of the call was to verify if the points suggested by the software for each article were appropriate and if the questions to be discussed in each of them made sense.
What was surprising to me was the accuracy of the topics/subtopics and of the questions identified. Out of some 30 questions (roughly 10 per each proposed articles) I found only 0ne that didn’t make any sense and two that required some polishing to make sense. Really impressive.
Finding the right content for a given audience (and its impact on a given distribution channel) is clearly crucial for effective communications and given the ever growing amount of content available the competition for eyeballs is stronger than ever. Tools like MarketMuse do just that.
In this case there is a team of technical writers that will create the articles, based on the topics that have been found by the AI software. There is, however, AI based software that is able to write the article on its own, like Eleuther.
Eleuther is an Open Source software, AI based, that can create coherent articles in good English, surely better than my Italish, once provided with a prompt, like the one resulting from the MarketMuse Suite. Take these two software together and you have an article generating application that will harvest content on the web and produce a brand new text.
Of course it remains to be seen what would be the value added in this activity. For sure, given the gazillion pieces of content on the web having a tool to make a syntheses and, even more interesting, creating continuous revisions that can point out advancement in a given area may be quite useful.
Would this software be able to provide insight? Not yet as far as I can see, but in the coming years…
Last year Microsoft fired dozens of journalists that were feeding articles to the MSN webpages, replacing them with AI. For sure AI is becoming a crucial tools for journalists, whether it will ultimately replace them remains to be seen. Most likely most news production will be AI assisted/generated by the end of this decade but a few most empowered journalists will remain to act as news orchestra directors.
Just yesterday I run into an article written by AI software. Take a look!