AI for human flourishing
Intelligence, guided by wisdom.
Reflections by Edwin Tan on AI, human agency, and the futures we might choose.
Opens in a new tab.
Personal reflections
Essays on AI, human agency, and the futures we might choose.
How I approach these reflections
My writings trace how my thinking has evolved amid accelerating breakthroughs. They’re an attempt to look not just at where we are, but where we might be headed, and to hold space for the quiet possibility that, if we get this right, something truly good could emerge.
These writings are not offered as prescriptions, but as invitations: born of curiosity, reflection, and deep humility. My thinking continues to evolve, and I hope you’ll approach these reflections with an open mind, unconstrained by today’s technologies or the limits of what we currently understand.
Consider Transformers, the neural architecture introduced in the 2017 paper Attention Is All You Need. By replacing sequential models like RNNs with self-attention and parallel processing, they radically expanded our capacity to scale language understanding. That single shift now underpins nearly all modern AI systems.
Then came AlphaEvolve, unveiled by DeepMind in May 2025, a general-purpose evolutionary coding agent built atop large language models. In a striking case of machine-led discovery, it autonomously designed an algorithm that broke a 56-year-old record in 4×4 matrix multiplication, reducing Strassen’s 49-multiplication benchmark to 48. No human had achieved that.
These moments remind us that we don’t have to be confined by today’s assumptions or mental models. My hope is to hold open space — for wonder, for moral imagination, for the possibility of something better — and to invite you, the reader, to do the same.
The OpenAI/Hugging Face Incident Reminds Me of the Paperclip Story
A decade-old thought experiment about a paperclip factory, and a 2026 incident report that reads like its lab notes.
From Mutually Assured Destruction to Mutually Assured Prosperity
An inquiry into our post-capitalist future
Beyond the Moving Goalpost
If AGI arrives on a Tuesday, will we even notice?
Guppies in a Tank. Humans in a System
How AI might preserve us, while erasing the freedom to shape our own world
The Founding Titan and AGI
A parable of power, sacrifice, and the true enemy
Why We Exist
AI in service of
human flourishing.
Eudavol AI exists to guide artificial intelligence toward a future that serves human flourishing. Our mission is to help steer AI toward what humanity would want if we were wiser, more united, and more deeply aligned with our highest values.
We call this guiding principle Eudaimonic Volition — the ideal will shaped by wisdom and a commitment to life.
Explore the ideas behind Eudavol
The Problem
Intelligence is accelerating.
Wisdom must keep pace.
Most AI development is driven by speed, competition, and profit, leaving little room to reflect on what truly matters.
This creates a dangerous gap: advanced AI is accelerating, yet the wisdom to guide it is under-resourced, undervalued, and increasingly urgent.
Why Now
Small shifts today.
All possible futures.
Even a modest chance of superintelligence in the next decade makes early action critical. And even if it comes later, what we build now will shape the future.
The stakes are too high to wait — small shifts today can echo across all possible futures.
Our Values
Eudaimonic Volition
Stewarding AI for flourishing futures.
Influence with Integrity
Aim to reach the few who shape the many.
Humility and Reflection
Hold strong values, not rigid answers.
Voices from the Frontier
“It’s incumbent on us to take this incredibly seriously — to have the best people in the world thinking about how to safely deploy these things.”
“My greatest fear is that… these digital beings we’re creating are just a better form of intelligence than people… We’d no longer be needed. If you want to know what it’s like not to be the apex intelligence, ask a chicken.”
“It’s my guess that powerful AI could at least 10x the rate of these discoveries, giving us the next 50–100 years of biological progress in 5–10 years.”
Our approach
The future of AI — hence humanity — may come down to the decisions of fewer than a hundred individuals.
We don’t aim to build frontier AI ourselves. We aim to influence the decision makers.
Partner with wisdom
Work with people who look beyond self-interest and hold a deep commitment to human flourishing.
Bring the right people together
Connect thoughtful people whose talent, perspective, and influence can shape the direction of AI.
Create space for reflection
Design environments where honest dialogue, shared understanding, and wiser choices can emerge.
Edwin Tan
Founder, Eudavol AI · Co-founder, CapBay
Edwin Tan is a father deeply concerned about the trajectory of artificial intelligence.
As co-founder of CapBay, one of Malaysia’s leading fintech platforms, Edwin built the mid and back-office, helping scale sustainable financial infrastructure and develop long-term systems thinking across Southeast Asia. He approaches AI alignment not just as a technical problem, but as a moral and existential challenge rooted in love for life and responsibility to future generations.
Email Edwin“If the future holds radical abundance, then wealth accumulation becomes irrelevant. If flourishing is possible, then our duty is to help unlock it — not just for ourselves, but for all who come after. This may become my magnum opus, regardless of the outcome.”
Why I started Eudavol AI
My journey began in 2016, when I first read The AI Revolution by Wait But Why. Since then, I’ve been closely following the trajectory of AI as someone deeply concerned about what we’re building and where it might lead.
One recent resource that’s helped me frame the broader possibility space is AI 2027. I understand these aren’t predictions or prophecies. They’re plausible trajectories. But I’ve been struck by how many of the dynamics described there seem to be playing out. Over time, I’ve noticed how consistently I underestimate the pace of progress, even when trying to account for exponential growth.
What concerns me most now is that the question of alignment — how we actually guide something this powerful — still seems unresolved.
I’m not an expert in safety or interpretability. I don’t have all the answers. But I do feel a growing responsibility to contribute in whatever way I can toward a wiser outcome.
That’s why I started this project.
Get in touch
Thoughts on an essay, or a question about Eudavol? I’d be glad to hear from you.