the build quality of the average house in america is actually insane...
somewhere in the mid 20th century, people just started accepting that they should live in a glorified chicken coop
We are going to kill Venmo. To start, we’re giving you $100 to send to your friends with AI.
Versine lets your AI send and receive money for you.
Works with ChatGPT, Claude, Muse, Instinct, or any AI you use.
Recipients don’t need a Versine account - they can cash out straight to their bank.
Comment and DM me your number, I’ll send you $10 on Versine: versine.com
is $80k a decent annual salary in Georgia, US for an SWE with 2 years of experience after university?
it’s hybrid (2 day/week in-office)
job security is guaranteed for at least 2 years (unless something extreme happens)
back when I was broke af, had no work experience and was getting paid pennies to do some repetitive online tasks, I often used to work 18+ hours a day whenever those tasks were available.
barely slept on some days.
they banned me cos they thought it wasn’t possible.
My workflow is pretty simple but it catches almost all issues. The code is always ready to ship, even though it requires some fixing to keep it future-proof.
plan + tips and worked examples first, by Opus. the plan is reviewed in a new session with fresh context. then I review it personally (if it’s an important or huge feature).
some rules, hooks and tests that are run before, during and after Claude edits something.
then an invariants reviewer, a convention reviewer.
then multiple adversarial reviewer agents that own and help maintain some specific critical modules.
every reviewer agent has all the domain specific knowledge required for its module.
At last agentic device QA with flows run by cheaper models that were initially created by an intelligent model.
then the PR is opened and I finally review critical parts of the code myself.
sometimes when I see people on twitter sharing AI slop code on twitter, I can't help but slightly doubt my workflow.
am I doing something really good that my AI code doesn't have such mistakes or am I too dumb to spot these mistakes?
I hope it's not the latter.
sometimes when I see people on twitter sharing AI slop code on twitter, I can't help but slightly doubt my workflow.
am I doing something really good that my AI code doesn't have such mistakes or am I too dumb to spot these mistakes?
I hope it's not the latter.
Homework, I'm taking a C++ class for my associates degree. I am doing pretty good so far but bombed the last quiz, the professor asks alot of questions in the book, even if you are a master at C++, you have to read the lecture slides or you'll bomb the quiz. The way they are worded is stupid.
in the modern world, you need to master one psychological component of your being:
you absolutely must be comfortable with deep levels of uncertainty at all times.
I always wanted to learn video editing and 3D animation but avoided because of just how tedious they are.
now that AI can handle all the tedious parts, it seems like the perfect time to get into video editing, vfx and 3D animation.
do you guys feel the same?
just asked Fable 5.2 to audit and fix my AI agents, skills, context files, etc. to improve the efficiency of my agentic coding and agentic QA workflows
So I woke up to this from @dodopayments. Out of blue they decided to close my account, no reasons given.
Sure, I am not earning but have loads of sign-ups for my app. That shouldn’t be a reason, or is it?
I was actually planning to slowly move other projects to Dodo, and make it the central payment processor. I guess I am never using Dodo again.
A feedback to the Dodo team, at least state the reason in the email for the user to know why this step was taken. Anyways, it was a decent experience using the platform. There were a few hiccups, but nothing major.
cc @ayushagarwal@goelrishabh