Silicon Valley’s biggest artificial intelligence (AI) developers, Scale AI and Appen, are looking for a new profile of data developer: poets. The sector aims to improve the creative writing of its digital text generators to give them a more human touch. That is why it needs writers, playwrights, novelists and humanities experts to feed its AI training systems with their own corrections and ideas.
Since the emergence of artificial intelligence and the popularization of chatbots like ChatGPT, many doubts have arisen about the consequences of its incorporation into our society. There is debate about whether artificial intelligence will be consolidated as a complementary tool that serves to facilitate routine processes in some jobs or if, on the contrary, it will replace workers in sectors as diverse as computer programming, graphic design or journalism.
There is no doubt that artificial intelligence is changing the paradigm in which companies work. The launch of ChatGPT, powered by OpenAI in November 2022, surprised many with its ability to answer complex questions and craft original-looking text. Since then, the developers of these artificial intelligence chatbots have been working to improve the quality of responses to user queries. However, doubts arise about whether this tool could replace human interaction in some tasks or jobs, since it has not yet managed to write with the sensitivity and literary rhythm characteristic of poets and creative writers or provide empathetic responses to users. .
But why this interest in improving the writing creativity of artificial intelligence? As Milagros Miceli, a researcher at the Distributed AI Research Institute (DAIR), explained to Rest of World, in recent months there has been a trend towards the professionalization of these chatbots trained with artificial intelligence. The developers intend to market this tool by offering their services to companies. For this reason, the need arises to diversify these chats and be able to cover the needs of the business market.
Thus, given the growing desire of companies to find systems that allow them to interact with customers without the human role, the need has arisen to “humanize” the responses of chatbots trained with AI to increase the quality of these interactions.
According to Gartner, it is expected that by 2025 around 85% of companies’ interactions with customers will be managed without the involvement of any human being. Furthermore, according to International Data Corporation (IDC), global investment in artificial intelligence is expected to reach $110 billion by 2024. This data points to a trend toward automating customer interactions, but there are still doubts. on whether these interactions will provide the same satisfaction to customers as contact with a human.
So why do they hire novelists and poets? Their role in this sector is to provide the companies that hire them with texts to form a database and train large language models or LLMs (Large Language Models). Hiring these profiles can be a way for companies to own literature and avoid copyright complaints, such as the lawsuits by George R. R. Martin, John Grisham and other writers against the creator of ChatGPT, OpenAI, for using their works to train these models of AI.
But training an artificial intelligence to generate literary language is not an easy task. Most of these LLMs are not trained to be creative, so they do not generate new expressions, but are limited to imitating human language that has already been published on the Internet or previously expressed by writers and poets in their works.
Fabricio Goes, professor of computer science at the University of Leicester, explained to Rest of World that artificial intelligence systems “are trained to reproduce. They are not designed to be great,” but “they try to be as close as possible to what already exists.” So, “by design, many people argue that those systems are not creative.”
Scale AI and Appen’s client rosters include some of the biggest names in AI development, including OpenAI, Meta, Google, and Microsoft. These companies are trying to take the lead on needs that AI has not yet met. There are still many languages ??in which this technology fails to provide a good service: an opportunity to cover this market and cover needs. For this reason, some developers are hiring highly specialized writers in languages ??such as Japanese and Polish.
Given this incorporation of artistic and literary profiles in the development of artificial intelligence, a question arises: is it a new job opportunity for writers or, on the contrary, are they throwing stones at their own roof?
Until now, chatbots have not been able to write with the sensitivity and literary richness that human poets and writers write with. This is the big difference between texts generated by AI and the human writing process. However, with the adhesion of poets and writers in the leading companies in the development of generative AI, it will be necessary to see if this technology will be able to improve to the point of being able to express feelings and perfect their creative writing, which is still very far from resembling human sensitivity. If it manages to improve this ability to express, will artificial intelligence then be a threat to writers and artists or will technology never be able to replace human creativity?
Anticipating lingering concerns about whether AI will be able to replace the work of human writers, a spokesperson for Scale AI said in a statement to Rest of World that its work developing generative AIs “has and always will include humans.” in the circuit, as it is essential to develop responsible, safe and accurate AI.” It only remains to be seen whether writers and poets will see this situation as a sustainable source of employment or a mere temporary job that will help AI eliminate the role of writers.
