ChatGPT: When AI Suddenly Becomes Truly Intelligent and Suitable for the Masses.

«It is difficult to imagine a world without IT, as it has become an integral part of our daily lives and the way we interact with each other and with the world around us. However, it is possible to imagine a world without IT in the sense that people could live, work and function without the use of computers and other electronic devices. Although the world would certainly look different without IT, it would not necessarily be a worse place...»
Ironically, this is the statement of a creation of modern IT. The author of the text passage is the intelligent bot ChatGPT from the laboratory of the American research organization OpenAI, which specializes in the development of artificial intelligence (AI). GPT is a language model that can generate eminently usable and grammatically correct texts. The program was developed using machine learning and trained with a large data set. It can be fine-tuned for specific tasks.

The chatbot improves itself - thanks to continuous learning

GPT has been trained with texts from books, articles, and websites and is able to generate text by predicting the next word in a sequence based on the preceding words. It does this by analyzing patterns and relationships in the training data and using this information to make predictions about which words are likely to come next. GPT can be used for a variety of tasks: including language translation, summarization, and text replacement. A GPT text bot can be tuned to perform very specific tasks: For example, creating image descriptions or answering questions. The perfect application area for this chatbot would therefore be the one we already know from various websites. However, ChatGPT is a quantum leap ahead: Thanks to AI, it can constantly expand its knowledge, so that eventually there absolutely will not be any questions it cannot answer in a certain area. ChatGPT can be tested on the OpenAI website[ii].

The AI future has begun

What ChatGPT can do is amazing. A short written request or instruction is enough, and the desired text, whether script, advertising text, slogan or newspaper article is generated (see below). What this means for the prospects of copywriters or journalists remains to be seen. In any case, there are experts who expect that even Google will soon have to deal with the enormous disruptive power of this development[iii]. After all, GPT delivers answers, not just search results - something Google has been striving to achieve for years, but has not quite managed to do yet. There is one big difference, however: GPT is not (yet) connected to the Internet. Its stored knowledge was programmed by humans.  Since it was only trained with data up to last year, the bot is not up to date today.

Incidentally, it is not only writers and editors who should be concerned with the creations of artificial intelligence. Graphic designers are also being challenged by bots. One of them, for example, is DALL-E[iv], a deep-learning model, also developed by OpenAI, that is able to generate a variety of images in a wide range of styles from text descriptions based on a dataset of text-image pairs - all within seconds. DALL-E works on the basis of a generative artificial intelligence model. Such models are able to generate new content based on data they have been trained with. 

«A tool designed to help people»

That people might feel threatened by bot competition is not surprising. But ChatGPT gives an excellently worded and reassuring answer to the question of whether its capabilities are a threat: «As an artificial intelligence, I do not pose a threat to humanity. My main task is to support users by providing information and answering questions to the best of my knowledge. I am not capable of acting on my own or engaging in harmful behavior. I am simply a tool that has been developed to help people»…

[i] Sam Altman

[ii] New Chat (OpenAI)

[iii] Hype um ChatGPT - Was kann der Chatbot wirklich?

[iv] Login Dall-E OpenAI