ChatGPT does not dream of electric sheep

Leicestershire, England, 1779. Ned Ludd reached his limit, couldn’t hold back any longer, and unleashed his fury on two of the newest modern gadgets: two mechanical knitting machines that made his presence in the factory irrelevant. Until he scrapped them, he couldn’t stop.

Decades later, this incident would give rise to a whole social movement, ludism, made up of those who opposed the introduction of machinery in the textile industry. They considered her a threat to their jobs and reduced their wages. To achieve their goals, the Luddites, inspired by Ned Ludd (probably a pseudonym), destroyed machines and burned factories, in a movement that was more proletarian than technophobic.

However, since then, anyone who opposes the new technological advances due to the economic and social losses that they could imply for the human being has been considered a Luddite.

The creative process no longer seems unique to humans

With the advent of artificial intelligence (AI) image-generating tools such as Dall-E 2, Midjourney, and Stable Diffusion in September 2022, the debate around the survival of artworks erupted into public opinion.

In December, a new type of tool, this time a text generator, sparked a new debate: ChatGPT was capable of producing texts in a conversational way in its interaction with the user, with astonishing ease and speed. So surprising is its use that some quickly tended to dust off (again) the well-worn Turing test to talk about the desired strong AI.

The debate extended to other areas of knowledge. In the educational world there was talk about how ChatGPT made old written homework unnecessary (there is already a tool that allows detecting jobs generated by ChatGPT). In the legal world, cumbersome jobs such as searching and summarizing case law can be a thing of the past. And these are just two examples. In each discipline there has been reflection, in public or in private, on the possible consequences of the introduction of artificial intelligence in the profession.

The need to measure the impact of ChatGPT

The attention that this tool is receiving is enormous, something unprecedented until now: in just two months, the platform that allows you to chat with ChatGPT reached a record figure of 100 million unique users (Instagram took 26 months to reach the same figure. Facebook, 54 months).

This expectation has also translated into numerous articles and news pieces in the media around the world and thousands of posts on social networks (some praising the tool, others mocking it or putting it to the test).

OpenAI had made a move, starting a race in which Microsoft and Google are in a hurry to enter to incorporate the new possibilities offered by this technology into their respective search engines.

News, books, papers, applications…

To monitor the impact of ChatGPT in our societies, a group of academic researchers have just launched a web project called ChatGPT Impact. It shows that since its launch on November 30, 2022, more than 21,000 news items have been published on ChatGPT in the media, more than 6 million tweets, more than 120,000 posts on Facebook and more than 17,000 on Instagram .

The impact is undeniable and its growth is exponential. In February alone, 35 new books a day about ChatGPT were published on Amazon. That same month, the number of scientific papers published doubled that of the previous month. There are more than 200 applications that use ChatGPT technology, while the content on YouTube, Twitch, TikTok, Discord and other platforms does not stop increasing. ChatGPT does not dream of electric sheep

These and other data will be collected periodically on the project’s website, with the aim of monitoring trends and analyzing the impact of ChatGPT on the different social manifestations.

We must not fall into the frustration that Ned Ludd felt when looking at these new “power looms.” Although AI surpasses us in certain capacities, we have to recover, as José Mª Lasalle said in a recent article, the meaning of tools and techniques: put them at the service of human creativity and imagination.

ChatGPT is very good at chaining words thanks to its statistical model, to such an extent that it looks like a person is speaking, but doesn’t understand what it is saying. He is not capable of reflecting or dreaming of electric sheep. We are the ones who can use that tool to imagine the future. We hope that our project helps to achieve this by showing the impact that it is having and, above all, that it will have on our lives.

No phrase generated by ChatGPT has been used in this text. In the near future, it may be mandatory to include this disclaimer in all the texts we read.

Fco. Javier Cantón Correa, Associate Researcher in the Department of Computer Science and Artificial Intelligence, University of Granada; Álvaro Cabezas Clavijo, Professor at the International University of La Rioja (UNIR); Pavel Sidorenko Bautista, Professor and researcher at the Faculty of Business and Communication, UNIR – International University of La Rioja and Rubén Alba Ruiz, Assistant scientist, Junta de Andalucía

This article was originally published on The Conversation. Read the original.