The Berggruen Institute is pleased to announce the 2025 winners of the Berggruen Prize Essay Competition, an annual award recognizing bold, original philosophical thinking that illuminates the most pressing questions of our time.<\/p>\n
The following English essay winner will receive a $50,000 USD award:<\/p>\n
Anil Seth, \u201cThe Mythology Of Conscious AI\u201d<\/p>\n
The following two Chinese essays winners will share a $50,000 USD award:<\/p>\n
Xin Huang, \u201cLanguage, Consciousness, and Computation: A Philosophical Analysis of the Token Concept in the Age of Intelligence\u201d\uff08\u8bed\u8a00\u3001\u610f\u8bc6\u4e0e\u8ba1\u7b97\u4e4b\u95f4\uff1a\u667a\u80fd\u65f6\u4ee3\u4e2dtoken \u6982\u5ff5\u7684\u54f2\u5b66\u5206\u6790\uff09Xiaoben Liu, \u201cThe First Paradigm of Consciousness Uploading: Mechanisms of Consciousness Evolution in the AI Axial Age and a Prospect toward Web4\u201d\uff08\u610f\u8bc6\u4e0a\u4f20\u7b2c\u4e00\u8303\u5f0f\u2014\u2014AI \u8f74\u5fc3\u65f6\u4ee3\u7684\u610f\u8bc6\u6f14\u5316\u673a\u5236\u4e0eWeb4 \u524d\u77bb\uff09<\/p>\n
This year, the jury selected one English-language winner and two Chinese-language winners. A total prize of USD 50,000 will be awarded in each language category, and all three winning essays will be published through Berggruen Press as part of the Institute\u2019s ongoing effort to amplify cross-cultural and cross-disciplinary ideas shaping the future.<\/p>\n
\u201cWe are very pleased to announce the winners of the second annual Berggruen Prize Essay Competition\u201d, said Dawn Nakagawa, the president of the Berggruen Institute. \u201cThe overwhelming submissions we received indicates that this initiative to find fresh voices and bold new ideas in this age of acceleration, is succeeding.\u201d<\/p>\n
This year\u2019s competition invited essays engaging with the theme of consciousness, intelligence, and the nature of mind in an age of advancing artificial systems, attracting approximately 3,000 submissions from over 120 countries worldwide in both English and Chinese. A rigorous blind review process was conducted to determine the winners, honorable mentions, and shortlisted essays.<\/p>\n
The Winning Essays\u201cThe Mythology Of Conscious AI\u201d
by Anil Seth<\/p>\n
In his provocative and incisive winning essay, Anil Seth, Professor of Cognitive and Computational Neuroscience and Director of the Centre for Consciousness Science at the University of Sussex, examines one of the most consequential questions of our technological era: Could artificial intelligence ever be conscious? While the prevailing intuition suggests that consciousness might naturally follow sophisticated computation, Seth outlines a series of arguments challenging this assumption.<\/p>\n
The jury praised Seth\u2019s essay as a powerful, sophisticated, and beautifully written critique of the assumptions surrounding conscious AI. Jenann Ismael (William H. Miller III Professor of Philosophy at Johns Hopkins University) described it as \u201cmeasured, intelligent, absorbing,\u201d noting how it compellingly challenges the widespread belief that advances in AI will illuminate the nature of consciousness, while highlighting the author\u2019s ability to weave familiar themes into a coherent and persuasive argument. Ned Block, Silver Professor at NYU, emphasized the essay\u2019s originality and depth, commending its exploration of four major ideas\u2014about computation, embodiment, the role of life, and the distinction between simulation and instantiation\u2014and calling it \u201coriginal, powerful and important.\u201d Robert Lawrence Kuhn, creator, host, writer of Closer To Truth, underscored the essay\u2019s piercing dismantling of our psychological biases and the hidden assumptions of computational functionalism, pointing out how it reveals the ways language and anthropomorphism seduce us into projecting consciousness onto AI. Together, the jury found the essay to be a thoughtful, pointed, and deeply insightful contribution to contemporary debates about mind and machine.<\/p>\n
Essay Abstract
Could AI be conscious? The answer to this question is of immense moral and existential significance. The prevailing view seems to be \u201cyes\u201d, on the assumption that everything that biological brains do can be captured by algorithms. There are, however, reasons to think otherwise. Our psychological biases predispose us towards projecting consciousness into our machine creations. The idea that computation is sufficient for consciousness rests shakily on a reification of the metaphor that the brain is a computer (it isn\u2019t). And there are many other possibilities, including that consciousness is property of life, not information processing. I critique the prospects for \u2018conscious AI\u2019, expose the distinct moral hazards posed by real artificial consciousness and conscious-seeming AI, and find a new relevance for an old idea \u2013 the soul.<\/p>\n
A note from Anil Seth, \u201cI was delighted when I learned that the topic for the 2025 Berggruen Essay Prize would be \u2018consciousness\u2019. Given my deep interest in consciousness from both scientific and philosophical perspectives, and my belief in the essay form as a powerful medium to condense thinking and reach broad audiences, I knew I had to write something. But I entered the competition with trepidation, knowing how popular the topic of consciousness would be. My essay focused on a question of deep societal importance: whether AI is, or could become, conscious. In particular, I wanted to give some reasons and resources to people who intuitively feel there\u2019s something wrong with the idea that AI might not only think but also feel, and to stake out a case that it might be life, rather than computation, that breathes fire into the equations of experience. I also wanted to draw attention to the ethical issues surrounding AI that is, or persuasively seems to be, conscious. I was enormously surprised, and thrilled, to discover that the essay won (for the English language prize). This means a lot to me personally, but I also hope it focuses broader attention on the issues raised in the essay. Whether you agree with me or not, simply catalysing a conversation will be a major win. I\u2019m also excited that so many people entered, and I hope that many new ideas about consciousness will have been sparked by this year\u2019s prize.\u201d<\/p>\n
Readers can now find Anil Seth\u2019s full essay published in Noema<\/a>.<\/p>\n \u201cLanguage, Consciousness, and Computation: A Philosophical Analysis of the Token Concept in the Age of Intelligence\u201d (\u8bed\u8a00\u3001\u610f\u8bc6\u4e0e\u8ba1\u7b97\u4e4b\u95f4\uff1a\u667a\u80fd\u65f6\u4ee3\u4e2dtoken \u6982\u5ff5\u7684\u54f2\u5b66\u5206\u6790) In this award-winning essay, sharp in perspective and distinctive in its point of entry, Xin Huang , uses the \u201ctoken\u201d to reopen the consciousness problem in the age of intelligent systems: when experience and intention are encoded as computable objects, do we obtain consciousness itself, or a more manageable substitute? The essay then examines the ambition of the \u201cunified token,\u201d an enticing blueprint for bridging language, computation, and mind, yet one that may ultimately amount to an engineering-level compromise and alignment. It further proposes the brain-computer interface (BCI) as a new route into consciousness research, suggesting that once intention, feeling, and action are translated into operable tokens, the problem of consciousness can be reframed and tested in new ways.<\/p>\n The jury highly praised Xin Huang\u2019s essay, recognizing it as a rigorous, original, and philosophically sophisticated analysis that elucidates the deep conceptual connections among language, consciousness, and computation in the age of artificial intelligence. Xinyi Fu, Professor of Philosophy at Fudan University, emphasized that the paper offers a clear and forceful distinction between the engineering-level functions of tokens and deeper semantic structures, convincingly arguing that the true significance of tokens lies not in constituting a unified world language, but in their role as dynamic and translatable interfaces. Jianhua Mei, Professor of Philosophy at Shanxi University, highlighted the essay\u2019s originality and theoretical depth, commending its construction of a unified \u201ctoken theory\u201d and its proposal of a novel criterion for assessing artificial consciousness based on token reconstruction and conceptual induction, which he regarded as a contribution of distinctive academic value. Zhihua Yao, Professor of Philosophy at the Chinese University of Hong Kong, described the paper as measured, insightful, and innovative, noting its development of the token concept into a general interface mechanism and emphasizing how its discussion of \u201cdistributed tokenization\u201d offers an illuminating perspective on artificial intelligence and brain\u2013computer interfaces. Taken together, the jury agreed that the essay makes a clear, profound, and critically constructive contribution to contemporary debates on language, consciousness, and intelligent machines.<\/p>\n Essay Abstract
by Xin Huang<\/p>\n
This paper takes the concept of token in Artificial Intelligence (AI) and Brain-Computer Interfaces (BCI) as a starting point to examine the relationship between language, consciousness, and computation. In AI, statistical modeling discretizes natural language into computable units, thereby exhibiting generative capacity; in BCI, continuous neural signals are segmented into representational tokens of intention, enabling externalization and parallel selection. Although differing in mechanism, both rely on tokens to compress complexity. However, this similarity remains confined to the engineering level rather than indicating semantic isomorphism. The meaning of tokens is always dependent on context and networks of difference, and their generation is shaped by algorithms, experimental conditions, and cultural settings. Accordingly, this paper clarifies why a \u201cunified token\u201d is untenable, emphasizing instead its value as an operational interface across systems; it proposes the autonomous generation and stable use of \u201cnew concept tokens\u201d as a criterion for machine consciousness; and it reveals how BCI, understood as a decentralized interface network, provides a new interpretation of mind-body interaction.<\/p>\n