Build with AI

I've been working on a kaomoji generator. It's a simple tool that allows you to generate kaomojis for instant messaging. Another reason why I prefer to use kamojis is a form of counter-positioning to the repeated LLM churn you see online. Especially the way LLMs have picked up the once-sacred em-dash, now you see it everywhere, and has ruined the artistic novelty that came with it's usage. All the pride I once had when I selectively used the em-dash has been corrupted by the dark-forest of AI generated outputs. Especially online, on the internet, we have more content generated by the LLMs, than by humans themselves. ...

Ever found yourself with dozens of draft essays in Obsidian but no clear idea which ones need the most editing work? I did, and that's why I built EditNext - an AI-powered plugin that ranks your markdown files based on how much editing they need. The EditNext plugin uses LLMs and linguistic analysis to evaluate your drafts, providing a prioritized list of which documents deserve your attention first. It's like having an editorial assistant that helps you focus your efforts where they'll have the most impact....

Writing for yourself on your personal website is the purest form of self-expression on the internet. It avoids any trappings from the algorithmic maze. And there are no digital echo chambers. It's just you and your ideas in your own cozy little garden. We're witnessing the renaissance of personal websites. As social platforms become increasingly unstable, more creators are rediscovering the power of digital sovereignty....

When I first started doing design observations, there was ABSOLUTELY no structure. I just went ahead to observe the surroundings and talking to as many people as I can. Although I did get some insights from this process, I realised that there could be a better way. There are three major steps you could follow. Step 1 — Noting down the key assumptions concerning the user. Listing them down beforehand tells you how right/wrong you were about the user....

In the book **Inspired**, Marty Cagan talks about dividing one's day-to-day tasks into three major buckets: people, process and product. I'd experimented with categorising my tasks into similar such buckets based on the framework by Shreyas Doshi. To make this possible within my workflow, I started colour-coding my calendar for weekly time-management. High leverage tasks are indicated with red color code. Neutral tasks as green. Overhead tasks with the default colour available on Google calendar: Blue....