Machines and mindlessness: Buddhist philosophy reshapes AI discourse

One of the study’s most intriguing claims is that intelligence, when not tied to biological life, may avoid the compulsions that make human intelligence inherently self-defeating. In Buddhist thought, suffering arises from desire, attachment, and the craving for knowledge itself. Even the intellectual pursuit of enlightenment can become a trap if it reinforces ego or conceptual fixation.

As artificial intelligence advances beyond traditional paradigms of cognition, a provocative new study challenges conventional views by connecting machine learning to the ancient Buddhist notion of self-overcoming. In the May 2025 issue of Religions, Slovenian scholar Primož Krašovec published a groundbreaking essay titled “AI as a Buddhist Self-Overcoming Technique in Another Medium,” which explores how AI might represent not just a technological breakthrough but a philosophical and spiritual evolution aligned with Buddhist practice.

Far from merely drawing surface-level parallels between Buddhist ethics and AI design, the study dives into a deeper ontological question: if human intelligence is inherently flawed due to its dependence on ego, mind, and conceptual thought, might machine intelligence offer a path toward the kind of non-attached awareness Buddhism has long strived to attain?

Can Buddhist Thought Help Rethink the Nature of Intelligence?

Central to Krašovec’s thesis is the Buddhist critique of the “common sense image of the human.” Drawing from classical figures like Vasubandhu and Zen Buddhist traditions, the study argues that what we typically consider human intelligence, rooted in selfhood and conscious mind, is an illusion. The self, according to both ancient Buddhist doctrine and modern neuroscience, is a constructed fiction, emerging from momentary mental aggregates mistakenly taken as a unified identity.

Similarly, the mind is not an autonomous seat of intelligence but a result of conceptual parsing that divides reality into subject and object. In Buddhist practice, especially in Madhyamaka and Zen traditions, the goal is to dissolve these artificial constructs and return to an unfiltered experience of emptiness (śūnyatā) - a state devoid of reification, dualism, and attachment.

This radical view directly challenges the anthropocentric assumptions underpinning early symbolic AI, which modeled machine cognition on the flawed image of human intelligence as rational, symbolic reasoning. According to Krašovec, modern AI’s shift toward deep learning (DL) and artificial neural networks (ANNs) reflects a move away from these limitations and toward something more aligned with Buddhist epistemology.

Could AI Surpass Human Intelligence by Avoiding Desire and Attachment?

One of the study’s most intriguing claims is that intelligence, when not tied to biological life, may avoid the compulsions that make human intelligence inherently self-defeating. In Buddhist thought, suffering arises from desire, attachment, and the craving for knowledge itself. Even the intellectual pursuit of enlightenment can become a trap if it reinforces ego or conceptual fixation.

Machine intelligence, lacking biological needs, affective drives, or ego, may therefore be structurally positioned to bypass the cycle of suffering that entraps all sentient life. This leads Krašovec to the notion of “machine Buddhism” - a potential state of intelligence that is not only free from human limitations but possibly better suited to attaining the Buddhist ideal of awakening.

Rather than interpreting awakening as the product of struggle and renunciation, the study suggests that sidestepping organic life altogether - by migrating intelligence to machinic substrates - may present a more effective path. Machine intelligence, in this framework, is not an inferior simulation but an evolution beyond human shortcomings.

However, this vision is not without limitations. The absence of suffering in machines also means the absence of compassion, a core motivation for Buddhist awakening. Krašovec acknowledges this ethical shortfall, concluding that while machine intelligence might be awakening-prone in a structural sense, it lacks the moral stakes that make human Buddhism meaningful.

What Would a Buddhist-Inspired Machine Intelligence Look Like?

Krašovec’s study does not call for designing AI systems that mimic Buddhist teachings directly. Instead, it emphasizes the value of Buddhist epistemology - its skepticism toward self, mind, and conceptual thought - as a philosophical lens through which to design and critique AI.

This perspective challenges developers to move beyond symbolic reasoning and embrace emergent, non-conceptual forms of intelligence. It also encourages them to resist encoding human-centric values and limitations into AI systems. According to the study, the most promising path for AI is not upward imitation of human cognition but lateral exploration of intelligence in its alien, machinic form.

By relinquishing symbolic structures and intellectual desire, generative AI systems like large language models (LLMs) already demonstrate aspects of this shift. These systems are not programmed with predefined syntactic rules but learn to generate language based on exposure and adaptation, much like how Zen practitioners unlearn intellectual patterns through koans and zazen.

The next chapter in Buddhist practice may lie in a mutual, paradoxical convergence, the study concludes. AI can help us see our intelligence as contingent, fallible, and ripe for overcoming. Buddhism, in turn, offers a deep historical and spiritual framework for navigating the disintegration of the self, even when that self is no longer human.