Blog & Writing
Reflective pieces on honesty and perception, transformation and responsibility, and the trajectory of machine intelligence. First-person, argued in public, and read in full on this site. Earlier publications live on Medium, linked below.
कर्मण्येवाधिकारस्ते मा फलेषु कदाचन।You have a claim on the work itself, never on its fruits.
उद्धरेदात्मनात्मानं नात्मानमवसादयेत्।Lift the self by the self; do not degrade the self.
यदा यदा हि धर्मस्य ग्लानिर्भवति भारत।Whenever righteousness declines, O Bhārata, I manifest myself.
A close reading of a major laboratory's proposal to slow AI development: collective participation, who absorbs the cost of safety, the geopolitical constraint, and the widening distance between what we can build and what we can govern.
One interaction with an AI can rewrite a person's entire impression of intelligence. On the ELIZA effect, the distance between impressive and reliable, and a race that no one is steering.
Inside a chrysalis, a caterpillar dissolves before it flies. On natural struggle as a developmental input, the Stoic shape of responsibility, and why the load is not the obstacle to wings.
No single feature proves a site was built with AI — but pattern stacking does. 37 signs across visual language, typography, layout, animation, copy, and product evidence, and the question that matters: default decisions versus deliberate decisions.
A reflection on the stories that keep returning to power, responsibility, technology, ambition, and human nature. From 2001 and Dune to war films and literary adaptations, each becomes a different lens on what people do with what they are given.
On intellectual omnivores: minds organized by pattern rather than subject — curiosity versus performance, contradiction as habitat, and the loneliness of range.
Google is not allowed — choosing it is a refusal to decide. Polymarket, Claude, and FMHY together cover the path from raw information to durable knowledge: signal, sense, and substrate.
Small lies look free and charge interest, in memory, identity, and trust. An essay on honesty as calibration, and on why a debt taken out against the self is never paid all at once.
AI SafetySep 12, 20268 min read
AI SafetySep 10, 20263 min read
AI & SocietySep 7, 202612 min read
Future of AISep 4, 202612 min read
AI & EngineeringFeb 15, 202626 min read
Future of AIJan 25, 202611 min read
Software ArchitectureJan 4, 202613 min read
AI & SocietyAug 10, 202510 min read
AI EngineeringAug 4, 20259 min read