Tech, Literature, Arts Experts to QNA: AI Will Not Kill Creativity

Despite the almost daily development in programs and applications of artificial intelligence (AI), it is impossible to abandon the human creator, experts in the fields of technology, literature and the arts have said.

In exclusive statements to Qatar News Agency , the experts said that the AI programs help create creative works such as writing stories, novels and poems, but it cannot match the capacity of human mind. It can help in preparing textual content of various types whether they are directed texts for social platforms, scientific research, articles, blogs, and many more, but it will not eliminate the presence of human creators.

With the accelerating pace of technological development in the field of AI and according to the latest studies, these programs have been supported by using text databases from the Internet, including books, magazines and Wikipedia texts, where 300 billion words have been entered into the system. One of the most popular AI programs like Chat GPT can on request provide text that appears as if it was written by a human writer. Perhaps one of the most important examples that confirm the success of AI programs in creative writing is what Canadian science fiction author Tim Boucher mentioned in the New York Post newspaper that he was able to write 97 small books in less than a year based on AI programs. This huge development in the techniques and production of AI programs prompted us to ask the question about the negative effects of such programs and their impact on creative or research writing as well as on design and various arts.

Dr. Ahmed Abdelali, senior software engineer at Qatar Computing Research Institute at Hamad Bin Khalifa University, told QNA that AI models by training computers with learning algorithms use data of very large sizes. This data contains all the content of the Internet in addition to many digitized books. These statistical models or that use of neural networks remember the graphic methods and linguistic structures and from that they save this information, adding that in the case of generating these models, the stored information is used and therefore these models cannot be described as intelligence.

Despite the ability of these models to remember as well as interact, they remain within the scope of what is known. Hence, the outputs of such algorithms cannot be described as creativity, he said, adding many may object to this and cite examples such as Portrait of Edmond de Belamy (2018) which sold for more than 432 thousand dollars. But this painting was the result of training a model of neural networks using more than 15,000 paintings between the fifteenth and twentieth centuries. In contrast to this, these algorithms can generate new ideas that the writer has never guess it before. With the help of machine learning and natural language processing algorithms, AI can analyze vast amounts of textual data to identify patterns and themes, where it can generate new ideas for stories and novels that a human author might not have thought of.

Abdelali continued that AI programs will remain the loyal assistant to researcher in scientific topics to help him arrange his ideas and output them in a consistent template. But researcher must carry out experiments and extract their results, and these programs can be used in the literary field to improve writing or speed up the process of writing and directing, he said, pointing to several technologies and models: (Alpaca (Stanford University), LLaMA (Meta)), GPT4All Nomic.ai, BLOOM BigScience, Cerebras-GPT Cerebras, GPT-J EleutherAI, indicating that these programs are open source, meaning that all the information that was used in training them is available to everyone and most of them use similar methods and may differ in the volume of data used as well as the training period, neural network architecture and the number of transactions used. (MORE)

Source: Qatar News Agency