Sherab Lhamo

Bhutan was recently ranked the top country to use ChatGPT, with approximately 15.96 percent of its population using the Artificial Intelligence (AI) tool, according to an analysis by Coinjournal.net.

Speaking at a session on  ‘Preservation of human intelligence in the age of AI’ yesterday at Bhutan Echoes:Drukyul Literature and Arts festival at the Royal University of Bhutan, the C0-founder of Curiouser.AI, Sonam Pelden, addressed the growing intersection of artificial intelligence and human intelligence.

Curiouser.AI, unlike ChatGPT, is an AI tool designed to enhance human capabilities by encouraging exploration through questions rather than providing direct answers.

During the session, Sonam Pelden stressed on the importance of human involvement in shaping AI and explained that the quality of AI’s output is intrinsically tied to the data it is trained on.

“Though it’s called artificial intelligence, it is created, built, and deployed by people,” she said. “Therefore, it matters who these people are and whether they bring a diverse range of perspectives and experiences to their work,” she said.

She said that AI, at its core, lacks inherent biases but can mimic its surroundings like a parrot. This issue is exacerbated by the dominance of white male engineers in the AI field, which can lead to the technology reflecting their biases.

“If they are the ones teaching AI how to think and understand the world, the technology they create will reflect their blind spots and their biases, therefore making the technology less effective and our society less equal,” she added.

Sonam Pelden also highlighted the issue of self-censorship in Bhutan, which has contributed to a perception that Bhutan is isolated from the global community. She challenged this notion by pointing out that Bhutanese people use American cell phones, drive Japanese cars, and engage with Chinese social media apps.

“If we really want to debate what is authentically Bhutanese, we must acknowledge that these elements are not traditionally Bhutanese either,” she said.

She argued that embracing modernity rather than clinging on outdated traditions is counterproductive.

During the session, she also demonstrated a biased Twitter thread response generated by ChatGPT on the topic of justice for Palestine and Israel. This example highlighted the potential for AI to spread disinformation and erode public trust in technology.

She shared that AI threatens free expression by supercharging the dissemination of disinformation leading to people losing trust in technology altogether.

Sonam Pelden highlighted that technology overlooks the needs of certain cultures and societies altogether. She attributed this to self-interest and a reluctance to abandon the views that they are familiar with, making it challenging to determine whose values should be prioritised in technology development.

She warned of the potential negative impact of AI on future generations, particularly the risk of diminishing cognitive abilities due to overreliance on tools like ChatGPT. This could hinder the development of critical thinking and problem-solving skills, she added. 

She emphasised the importance of maintaining human control over AI, if not, it could lead to technology shaping human identity and values. “And if this current trend continues, we will soon reach a point where humans will humbly wait for an algorithm to tell them what to do and how to think,” she said.

She said that preserving Bhutanese identity in the age of AI requires maintaining unique human qualities by ensuring that AI systems, which are trained on diverse data, can recognise and appreciate multiple forms of beauty and value.

“AI is designed to create standardised content with a digital landscape filled with repetitive information and overreliance on recycled data has transformed the internet into an echo chamber,” she explained, “It is a classic case of garbage in, garbage out.”