Carson @the_deep_coder

Don't reinvent the wheel. If a feature for your product already exists in open source, copy it and tweak rather than building from scratch.

Want to build a billion dollar startup? Get really really good at one thing first.

When vibecoding, start by taking your raw idea and refining it. I like having GPT generate a PRD (Product Requirements Document) that gives me more specificity while prompting

Whats your reason for putting in so many hours into your startup?

I think product managers are the most successful vibecoders. They understand the level of specificity required to give instructions. They also know how to plan a project, understand requirements, and handle edge cases. Want to get good at vibecoding? Study PM.

What percent of your posts are product updates/milestones?

How much money does it take to retire and start a chicken farm? Asking for a friend

X is actually so cool. You can document your whole journey for the world to see. Then you'll look back one day and say "Wow, those sleepless nights were really worth it!"

As a first-time founder, it's important to experiment as much as you can. Learn what works and what doesn't. Note it down so you don't forget what doesn't work. I promise if you're disciplined about experimentation, you'll reach a new level.

GPT kinda deserves equity in everyone's company

Ngl this week's been hard. Got an 81 on a test I studied hard for because I bubbled-in the wrong answer. Rejected from an internship after an interview and referral. Slow growth on X. But I've said consistency is the most important thing. Now it's time to stay true to my word.

If you're building an app, look into emotion-based design. Quick animations, milestone-based structure to incentivize, and enticing colors. It's what apps like Duolingo and Headspace used to skyrocket consumer retention.

Being a software engineer means signing up for a career of continuous learning.

Building a startup can require intense hours, working without results, and sharp deadlines. How do you guys protect your mental health while shipping?

Beginning of product: "Ok let's plan this in figma, design with a modular architecture, plan out every prompt with gpt beforehand..." Middle of project:"If GPT doesn't generate this properly I'll tell it the code is wrong and ask it to regenerate" Relatable?

Don't build features. Build solutions. New features not only take more time, they also increase app complexity. Growing requires shipping fast. Make sure your efforts lead to the greatest rewards. Focus on building what your USERS need instead if what YOU want.

Day 3 of building FairDocs: AI-Based Participation Grader ✅ I can now get the revision summary of the contributors. It has a few issues but the base idea is there. Next: analyzing revisions and determining weightage scores to give each user based on them.

Scientifically, you should soldier through. The brain hates unpredictability; new tasks or stressful situations produce high prediction error. With repetition, prediction error lowers, and the stress regions are less active. Pushing through becomes easier the more you do it.