Vivek Haldar

Freshly inked

How I Cut an AI Agent's Token Use by 94%

I’ve been writing about specialized agent harnesses and compiling them from task specifications for a while, and I realize the idea can sound abstract. So here’s a concrete example from a workflow I run every day.

Recent essays

Jul 4, 2026 · AI & LLMs

Cognitive Debt and the Future of Programming

Geoffrey Litt recently gave a talk at AI Engineer called “Understanding is the new bottleneck”. Simon Willison boiled its central phrase down to “understand to participate”: if agents are writing more and more of our code, we still need enough understanding of …

May 29, 2026 · AI & LLMs

The tension between local and cloud agents

I’ve grown to strongly prefer agents running locally on my own machine. I was doing so even before the latest crop of features from Anthropic and OpenAI that made it possible to remote-control local agents from their mobile apps — something that has made the …

May 21, 2026 · AI & LLMs

From Emacs to Agents

I used to be a heavy Emacs user. Note the past tense.

May 15, 2026 · AI & LLMs

The Complexity Ratchet

Recently, Jasmine Sun wrote a long op-ed in the New York Times about the notion of a permanent underclass. Everyone at the big AI labs is convinced that AI will eventually replace all human labor — all human knowledge work — and that if we don’t calibrate our …

Apr 23, 2026 · AI & LLMs

O Moat, Where Art Thou?

How does one build a moat in the AI startup space? Or as VCs like to say: “Where does the value accrue?”

Everything else, year by year

2019 1 essay
2017 1 essay
2016 2 essays
2015 1 essay
2013 74 essays
2012 86 essays
2011 48 essays