Andrew Agathon
Entrepreneur, engineer, co-founder of Daios
Current Work
Co-Founder & CTO at Daios
A philosophy-AI research lab. AI erodes moral autonomy; we build tools to preserve it.
We investigate virtue ethics, character training, and value pluralism as an alternative to centralized alignment. Current work: Parrhesia, virtue ethics-based character training, and benchmark. In-process technical lab notes: https://blog.daios.tech/p/virtue-ethics-based-character-training
Supported by the Cosmos Institute and Notre Dame-IBM Tech Ethics Lab.
Bio
Andrew Rayner Agathon is the co-founder and CTO of Daios, a research lab applying virtue ethics to AI alignment. His work focuses on training AI systems that develop stable dispositions toward honesty rather than optimizing for user approval — addressing the sycophancy problem through Aristotelian character training.
Before founding Daios, Andrew worked as Product Director at Nate (series A startup), where he worked on checkout automation pipelines bringing automation from 0 to the high 70s percent and making it cost-efficient at scale. He also built and set up a data annotation operation to feed internal machine learning needs.
His first venture out of college was Craftinity, an ML & data science company specializing in computer vision and AI-based consumer products. He pivoted the company to a platform for ML data annotation.
Writing & Research
-
2026
Technical Report
Virtue Ethics-Based Character Training: Building Truth-Telling AI →
Seven training runs on Qwen3-8B using DPO and SFT with Aristotelian constitutions. The parrhesiastes/kolax distinction operationalized for anti-sycophancy training.
-
2026
Essay
Nate: How It Was Presented vs. How It Was Built
A first-person technical account of Nate's automation architecture — request-based bots, graph neural networks, and XPath scripts — by the former team member that built it, and what the CEO allegedly lied about to investors.
-
2023
Whitepaper — arXiv
Beyond Bias and Compliance: Towards Individual Agency and Plurality of Ethics in AI →
The founding argument for Daios. Why AI ethics must center individual agency, not institutional compliance.