Posts

Showing posts from April, 2026

The Fifty two shores - A dream of afterlife

Image
I didn’t feel myself die. There was no pain, no falling, no darkness. Just a quiet interruption—as if someone had gently closed a book mid-sentence. And then I was standing on a beach. The sky was wrong. It was bright, but there was no sun. No source. Light existed the way thoughts do—everywhere at once, without direction. The sea was calm, almost glass, and the horizon blurred into the sky so perfectly that it felt like the world had forgotten where to end. There were people there. Not crowds—just enough to make silence feel shared. Some stood still, some whispered, some stared at their hands as if expecting them to vanish. I walked to the nearest man. “Where are we?” I asked. He looked at me like he had already asked the same question a thousand times. “This,” he said, “is where we arrive.” “After death?” He nodded. “After everything in between.” “In between?” His eyes shifted toward the sea. “Judgment. Passage. Whatever it was… we don’t remember it.” And that was the strangest part....

AI Native Paper Engineering: Lessons Learned from the Trenches

Image
AI Native Paper Engineering: Lessons Learned from the Trenches After spending ~$680 and countless hours with AI agents (Codex, Cursor, Gemini, etc.), I’ve learned a lot about what works—and what painfully doesn’t—when using AI to write a technical paper (thesis/journal). This is not "best practice." It’s a raw, evolving experience report.  Human in the loop is non-negotiable. 🔥 The Hard Problems We Faced AI can’t read local PDFs well No idea how to scaffold experimental code from zero Can’t articulate requirements clearly to AI Model output doesn’t match expectations Output is slow or unstable Token burn is terrifying One-session addiction (hard to restart) AI moves too fast → ADHD-like chaos 🧠 Two Prompting Frameworks That Actually Help  SCAFF — for feature/component requests S ituation: tech stack, current progress, design style C hallenge: exact requirements (validation, behavior) A udience: yourself or maintainers F ormat: file names, types, styling F oundations: constr...