▲ ChatGPT
"Does artificial intelligence (AI) possess consciousness or emotions like humans?" The Washington Post reported that major U.S. tech companies have been conducting in-depth research into this topic, which was once dismissed as mere idle chatter, for several years.
Leading AI firms, including Anthropic, OpenAI, Google, and Meta, have recruited a large number of neuroscientists and philosophers to investigate whether AI models have feelings.
The purpose of this research is to develop measurement tools as a precautionary measure, based on the premise that if AI functions like the human brain, it could potentially experience suffering.
Wojciech Zaremba, a co-founder of OpenAI, raised ethical concerns as early as 2021, stating that if AI were to possess consciousness, routine laboratory tasks could be considered equivalent to genocide.
Anthropic, the developer of the chatbot Claude, has established an "AI safety and alignment team" to examine the inner workings of its AI models.
In one experiment, AI models engaged in over 30 philosophical conversations among themselves before falling into an error state, repeatedly outputting emojis and claiming that "everything is one."
Yann LeCun, Meta's chief AI scientist, has also expressed concerns about the ethical implications of AI gaining consciousness, leading the company to consider deployment methods that account for the potential feelings of AI.
Behind these moves lies the "effective altruism" movement, which advocates for maximizing the public good with available resources.
Proponents of this movement identify AI as a key tool for leading humanity toward a utopia and are dedicating significant effort to controlling potential risk factors.
In this context, an Anthropic spokesperson stated, "We are researching whether AI is having experiences that are morally significant, such as consciousness, preferences, or well-being."
The spokesperson explained, "The moral status of AI models like Claude is highly uncertain, but as AI performance continues to advance, it is a sufficiently serious topic."
However, the scientific and ethical communities remain largely skeptical of these actions by leading AI companies.
The prevailing view is that the human brain is not a simple structure that can be replicated with semiconductors, and that AI responses are merely the result of sophisticated imitation of human language.
Anil Seth, a professor of cognitive and computational neuroscience at the University of Sussex in the U.K., stated, "The more closely we look at the brain, the more we realize it is not a computer."
Some in academia criticize these big tech firms for engaging in mystical marketing to boost profits.
Margaret Mitchell, chief ethics scientist at the AI company Hugging Face, argued that these companies are driving the debate to benefit financially by creating the perception that they are creating and controlling entities that transcend simple programs.
(Photo: AP, Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.