4 Ways I’m Thinking About AI From a Buddhist Perspective

A personal take from Ryan Stagg

In our last two posts, we shared a field framework and a qualitative synthesis capturing how Buddhists are thinking about AI. As our initiative moves forward, we plan to share more of this institutional work, but also less formal posts from our team and community members like this one, which was written by team member Ryan Stagg.

---

1. In Search of Coherence

For the last two years, I’ve dedicated the majority of my professional energy to understanding and exploring AI. The more I learn, the more I feel like I’ve barely skimmed the surface. There’s something inherently overwhelming about the phenomenon of AI. Every day, machines are generating and processing information at scales far beyond what humans could possibly read or verify. New models seem to arrive every week. They’re so good at creating text, images, and videos that it will soon be very difficult to determine the origin of digital information (not that it’s all that clear to begin with). Prompt an LLM with a simple sentence and a second later you have a five-page treatise. AI tools are being broadly adopted in seemingly every sector of society, and are increasingly shaping the media through which we come to understand our world. Our faculties of attention are under daily mimetic barrage. It’s like AI is racing out ahead of us, driving the world we used to know further into the unprecedented and uncertain.

While fully acknowledging the real-world benefits and incredible access to information that AI affords, I think the daily habits needed to keep oneself oriented rather than overloaded are becoming harder to maintain. Our nervous systems and sense-making capacities are maxing out amid the velocity and abundance of content we’re all having to navigate on a daily basis.

Given all this, I’ve been thinking a lot about coherence, particularly as a quality of mind. As social media pours out snippets and fragments of information devoid of context and narrative, and LLMs stochastically process and reflect our thinking, the fractures grow. Humanity’s existing institutions and narratives are epically failing to hold the pieces together.

I think it’s fair to interpret Buddhist practice as the intentional cultivation of coherence. By acting morally, you refrain from internal and social divisions (when you lie, for example, you splinter off a false narrative and experience the stress of maintaining it). In meditation, you steadily gather the strands of mind together. From this coherent base, insight can arise. There’s some scientific backing for this interpretation in that meditation induces states of “neural synchrony” in the brain. The ability to generate stability and coherence of mind through one’s own practice is becoming an all-too-necessary life skill. It’s like generating a signal amid a deluge of noise.

As relevant as Buddhist teachings are in this moment, the speed and scale of AI has put us in a place where few contemplative teachers are attuned to the unique textures of our now AI-mediated world. We need new articulations and new language that capture the subjective experience of Buddhist insights in this context. How are greed, aversion, and confusion manifesting in the race to superintelligence? How do we skillfully navigate a world increasingly shaped by machines? How do we orient ourselves to wisdom and compassion in virtual reality? How do we find coherence amidst so much noise? We need a contemplative guide for navigating the disorienting experience of living in the age of AI, and I hope our work can help inform that.

2. Nature, Embodiment, and Home

When I’m overwhelmed and disoriented, I find it helpful to go sit on my deck. A Green Ash tree hangs overhead, and squirrels, bluejays, cooper’s hawks, and even bald eagles(!) occasion the view. After a little while, I can settle into my body and my breath. Throughout our conversations about AI with Buddhist leaders, we’ve repeatedly heard a call to return to our bodies, communities, nature, the land. Anecdotally speaking, most of my Buddhist friends are somewhere between “I sold everything and started a farm” and “I think I overdid it on the houseplants.” I’m trying to figure out what to make of this agrarian attraction.

The view from my deck in Colorado (it’s a shared space).

One of the major currents in modern Buddhist teachings is embodiment. There are clearly grounds for this in foundational Buddhist teachings, but I think the way it’s being taught now is a unique response to the digital age. I’m sure you’ve noticed that when you stare into screens you can lose track of your physical body, your surroundings, time, and thought. Estimates of average screen time across computers, smartphones, TVs, etc., range from 7-12 hours a day for the average person, meaning many of us spend the majority of our waking life “digitally immersed.” That contemplative teachers are emphasizing the importance of inhabiting the body as a counter to digital immersion seems skillful.

In addition to this longing for nature and to be somehow more embodied, there’s a discourse emerging in Buddhist circles around digital vs. analog. Digital is discrete 0s and 1s, on and off. It’s quantized bits that can be perfectly replicated. Analog is a full wave that changes continuously over time. It’s context-rich, capturing a provenance that fades over time. There’s background noise, it’s messier, and entangled. While our brains are quite capable of operating in digital environments, they are analog.

While there’s some danger here in grasping at an agrarian, minimalist, retro fantasy world rather than working with the one we’re in, I think these inclinations are all signs of our basic wisdom at work: seeking center, seeking home, seeking groundedness. Renunciation is a key practice in Buddhism. It’s the beginning of the path, where we give up seeking fulfillment in unfulfilling things. While it looks different across Buddhist traditions, there’s an underlying discrimination between what is conducive to contemplative practice and what is not. I’m not suggesting we all take up a monastic life with only a robe and an alms bowl, but I think our use of AI and digital technology in general should be very deliberate. I’m not sure enough can be said about simply foregoing screen time for time spent attending to the felt sense of the body in exercise, nature, books, face-to-face conversation, and vinyl records.

3. Cognitive Debt vs. Augmentation

AI systems are reshaping our minds. How we think. How we pay attention. How we feel. The delineation between human and computer is getting murky. This is obvious from a Buddhist POV, where all phenomena arise interdependently from causes and conditions. It’s also obvious in contemplative science, where this insight finds kinship with neuroplasticity and mental training. But we know precious little about how our minds are changing in response to AI. A recent study by MIT Media Lab suggested that using AI to write essays results in “cognitive debt.” Basically, in the frictionless ease of using AI, your brain atrophies. Many technologists are positioning AI as a “co-pilot” and a way to augment one’s cognition. One small study isn’t enough to draw final conclusions, but it makes sense to me that the default path of AI use is not cognitive augmentation but cognitive debt. Intentional effort and mental exertion are the currency of mental training, neuroplasticity, and cognitive development.

Many Buddhist thinkers are building their analysis around the potential consciousness/sentience/awareness/subjectivity/selfhood of AI systems, which is all very interesting. I have my thoughts about that, but I’m more focused on the downstream consequences of what one believes. I recently found a study from Rose Guingrich, a postdoc at Princeton, who is studying beliefs about AI consciousness. Apparently, if you think AI is conscious, then the quality of that interaction will shape how you treat people later, and if you don’t think it’s conscious, it doesn’t. It seems one’s belief of consciousness raises the interaction into the social systems of one’s mind. Good to know.

These studies are just the tiniest tip of the iceberg. Richie Davidson once described our relationship with technology to me as something along the lines of “the biggest experiment in mind training ever.” It would be wise to know a whole lot more about how AI is reshaping our minds, and contemplative neuroscientists in particular are well-positioned to help us here.

4. Buddhist Ethics and the Co-Dependent Origination of AI

The bhavachakra (wheel of life). The twelve links of dependent origination are depicted on the outer edge of the wheel.

The project I’m most excited about is bringing frameworks from Buddhist ethics to practical AI alignment discussions. It’s a long bridge to cross and we’ll need better scholars than me to make it work, but perhaps I can help get the conversation started. As part of this effort, I’ve drafted a “causal chain” of AI with inspiration from Kate Crawford’s Anatomy of an AI System, which sheds light on the invisible processes behind AI, and Buddhism’s 12 Links of Co-Dependent Origination. My general intention is to greatly expand the scope of the alignment discussion, which is too narrowly focused on outputs and benchmarks.

1. Intentions and Worldviews: The motivations of developers, organizations, and markets that shape why AI systems are built

2. Data: The remnants of human activity and substrate of machine learning

3. Datasets: Curated, filtered, and assembled bodies of data that form the basis on which models are trained

4. Training: Optimization processes that engrain patterns in the model’s weights

5. Post-Training: The focus of most AI ethics and safety work. Aligning outputs to social norms.

6. Deployment: The user interface, persona, and monetization strategy through which an AI model is made available

7. Inputs: User prompts and memory

8. The Black Box: The model’s internal process bridging inputs with outputs

9. Outputs: The model’s visible actions (text, image, voice, video)

10. Psychological Impacts: The attentive, affective, cognitive, and behavioral effects on human minds from interacting with AI systems

11. Social Impacts: Effects on communities, institutions, and governments

12. Environmental Impacts: The ecological externalities of the entire AI pipeline

13. Feedback Loops: Numerous feedback loops exist throughout the chain

At each link in the chain, we can think of ways to cultivate the wholesome and down-regulate the unwholesome factors at play. Take “data” for example, it may seem like a neutral link in the chain, but there are serious ethical issues around digital surveillance, privacy, consent, intellectual property, and copyright infringement.

In my opinion, current AI alignment strategies risk becoming performative, where outputs appear aligned but underlying intentions, behaviors, and real-world impacts are not. Rather than focusing only on discrete metrics and benchmarks, we need to be thinking in terms of a multidimensional alignment ecology constituted by continuous relational feedback loops between AI systems, humans, and the world. My colleague Peter Herschock has offered the framing of attunement instead of alignment to describe such an interdisciplinary, entangled, and dialogical approach.