Exhibition

at Quantum Art Festival Quantum Art Festival 2/4

https://www.artfesq.com/

Dates: 7 (Thu) - 9 (Sat) December

Venue: LIGHT BOX ATELIER 2F (6-13-1 Minami Aoyama, Minato-ku, Tokyo) Opening hours: 11:00 ~ 18:00 Free admission

ABSTRACT

Presenting intentionally extreme and radical AI-generated content to ‘advice' to human society

The ARTIFICIAL ADVISORY project is a series of artistic explorations of how artificial intelligence might rewrite the future, and how human intelligence can overwrite or co-write it in a humanistic way. The project focuses on exploring the realms of human authenticity and autonomy in the age of generative AI through artistic experimentation and journalistic investigation.

The ARTIFICIAL ADVISORY project, is a Turing test-like exploration using complex generated AI content. As an artefact, I published a futuristic magazine called BASH, which mimics futuristic technology magazines like WIRED. BASH magazine describes the future of quantum computing in 2040, with all text and photos generated by AI and printed in a 60-page magazine with almost no text editing (only cutting and pasting to adjust the number of characters to fit).

The pros and cons of using generative AI for research are currently being debated in the field of science, the centre of human intelligence.

New research is being generated every day in generative AI. The latest research reported on the arXiv, the world's largest pre-print server, shows that Large Language Models (LLMs), such as OpenAI's ChatGPT, are used in peer-review of articles (i.e. attempts to assess whether the content is suitable for publication in an academic journal). *1).

Meanwhile, a survey of more than 1,600 researchers by the British scientific journal Nature (*2) on their attitudes to AI revealed that researchers are caught between their hopes and fears about AI. More than half of the researchers expected AI tools to be 'very important' or 'essential' for use in their respective research fields over the next decade. The researchers who responded to the survey also identified ChatGPT and LLM as the most impressive and useful AI tools in science. However, in terms of the application of generative AI to science, the survey revealed that 68% were concerned about the spread of misinformation, 68% thought that plagiarism would become easier and more difficult to detect, and 66% were worried about inaccurate information being introduced into research papers.

Will Science continue to be a product of human intelligence?
Or will it be the product of human and artificial intelligence?
Whose future will emerge from it?
How will the future be found in the first place - and how will it be found?

In response to this situation, the art project Artificial Advisory explores the question of whether humans can imagine and create a future that is inherently different from AI, by daring to use explicit generative AI representations. Using ChatGPT and the image-generating AI Midjourney, to question the viewer on the possibilities of human intelligence and, at the same time, its impossibility.

The Quantum Art Festival uses a quantum computer as an object of artwork, attempting to make "advisory" to human intelligence.

*1 Can large language models provide useful feedback on research papers? A large-scale empirical analysis https://arxiv.org/abs/2310.01783 *2 AI and science: what 1,600 researchers think https://www.nature.com/articles/d41586-023-02980-0

PROMPTS

NEUROCOOL:COOLING THE FUTURE

In 2040, humanity is on the brink of holding the rise in global average temperature below 2 degrees Celsius. This prompts the emergence of an anthropocentric, short-term device that lowers the temperature of only humans by 2 degrees Celsius. As research, we asked the Chat GPT to think of ways to adapt the Selective Brain Cooling system used by Thomson Gazelles and others in the desert to humans. The name NeuroCool and the use of microfluidic devices are both suggested by Chat GPT. Incidentally, the research here has also led to the development of a new method for the use of microfluidic devices.

Their research on 'localised ondemand analgesia' (*), in which microfluidic devices cool peripheral nerves and remove pain, was reported in SCIENCE magazine. It is unclear whether ChatGPT refers to this.

The Chat GPT has also produced a valid argument that NeuroCool is "not a fundamental solution" to global warming.

*Jonathan T. Reeder et al. ,Soft, bioresorbable coolers for reversible conduction blocks of pe- rior nerves.Science377,109-115(2022). - DOI:10.1126/science.abl8532

REIMAGINE LIGHT

In order to chart the progress of 'quantum biology' in the year 2040, Chat GPT studied a paper (*) with leading researcher Jim Al-Khalili as author, and predicted the innovations that would occur in the year 2040.

The results of the study were used to predict the innovations that will occur in the future in 2040. Among these predictions was artificial photosynthesis, for which Chat GPT generated 'an article written by a fictional science journalist', giving rise to the supposedly new name 'quantum biocell'. As a hypothetical application, it also mentions the possibility of quantum biocells being incorporated into biocompatible interfaces that could be implanted in or attached to human skin. Scientific integrity aside, there is much that is suggestive.

* Kim, Y.; Bertagna, F.; D’Souza, E.M.; Heyes, D.J.; Johannissen, L.O.; Nery, E.T.; Pantelias, A.; San- chez-Pedreño Jimenez, A.; Slocombe, L.; Spencer, M.G.; et al. Quantum Biology: An Update and Perspective. Quantum Rep. 2021, 3, 80-126. https://doi.org/10.3390/ quantum3010006

FROZEN TIME IN THE WAR

A setting was created in which 'long-life medicine' will be created in 2040, which will enable people to live longer. When such a medical system is created, it is thought that there will be a situation in which 'the best allocation is not given to those who need it most'. As the most appropriate personality to represent such a critical event, we created the character Elara, a climate refugee woman born as a result of the food crisis related to climate change. The author then asked ChatGPT to analyse the literary work “War's Unwomanly Face” (by Svetlana Alexievich), which depicts women in the aftermath of war, and to take on the persona of the climate refugee woman and chat with the author. The students are asked to take on the role of a woman, a climate refugee, and chat with the author. The generated chat was turned into a question-and-answer interview.

Prompt given to the ChatGPT as a characterisation:

I am a woman who has been living in a refugee refugee camp. In the future, in the year 2040, when the 'longevity certification system' has been established, a woman named Elara has come for an assessment of her eligibility to receive government longevity care. She is a refugee from war-torn Europe who has applied for longevity in order to find her own daughter, who was lost in the war. However, her application is rejected for unreasonable reasons.