Software, AI, and side projects. Learning, building, and sharing. Former Heroes World Champion • Ex @TeamLiquid & @Cloud9.
Bay Area, Norcal
Joined October 2014
- Tweets3.3K
- Following316
- Followers5.6K
- Likes7.1K
Pinned Tweet
I used to compete in esports with Team Liquid & Cloud9.
Now I build software and AI tools, mostly by breaking things first and figuring them out second.
No real plan, just a lot of testing, iterating, and occasionally wondering why something stopped working.
Git and side-projects coming soon
Despite my Ex Wife’s attempts to drag the process ad infinitum, I found a way to bifurcate my married status, with the judges blessing, effective immediately. 😎
Never give up no matter what they throw at you. Stay true to your goals and keep pressing forward 🤙
My "find people nearby" function was way too slow in a proximity based social app side project I am working on, so I built a tool that got that exact function running about 50x faster.
It's called LoopForge. It tries faster versions of your code, checks every one against the original, and throws out anything that changes the result. You only keep what's faster AND still correct.
I uploaded the tool for others to try it as well. It includes a demo using my findings as an example: It will walk you through the process as it tried to speed up my code, attempting multiple avenues and eventually showcasing the two faster versions that broke, and then the ~50x one.
Free and open source (MIT). About 30 seconds to try, steps in the reply.github.com/johnpaullopez13-s…
Try it yourself (no setup, no signups, no API keys):
git clone github.com/johnpaullopez13-s…
cd loopforge
pip install numpy
python run_demo.py
You'll watch it test four versions of a function, drop the two that break it, and keep the fast one that still works.
A bit more for anyone who wants to run it on their own code:
You can give it your function, a few example inputs, and a way to check the answer is right. It tries new versions and only keeps the ones that are faster and still correct.
The demo uses pre-written versions so you can watch it with zero setup. The real idea is to let an AI write the new versions for you. The instructions it follows are in the repo (AGENT_PROMPT.md), so you can plug in your own model.
It works best on small, self-contained code like sorting, matching, scoring, or number crunching. Not for code that mostly waits on a database or the internet, since the slow part there isn't really your code.
AI isn’t replacing us, it’s refocusing us on what makes us human
The most common question I hear about AI is:
“Is it going to take my job?”
But that question misses the deeper point. A more thoughtful inquiry might be:
“Which parts of my job were never really meant for a human in the first place?”
For generations, we’ve spent enormous mental energy on tasks that don’t require imagination, empathy, or judgment. Things like sorting emails, copy-pasting data, writing status reports that no one reads.
These aren’t jobs, they’re chores disguised as work.
AI, for all its complexity, is remarkably good at the dull stuff. And that’s good news. Because it means we can return to the kind of work that actually requires a brain.
A marketing coordinator shifts from recording numbers to shaping narratives.
A customer support rep stops repeating themselves and starts improving the system.
An ops manager stops chasing data and starts designing better flows.
AI isn't meant to replace people, it nudges us from reaction to reflection.
And the greatest thing it gives us?
Time.
Time to think more deeply.
Time to connect more meaningfully.
Time to do the kind of work that makes us feel human.
So instead of fearing what AI might take away, let’s get excited about what it’s handing back.
The freedom to focus on what truly matters.
Welp it was fun while it lasted. GGs all
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
John Paul "KingCaffeine" Lopez ⚒️ retweeted
me as i read 40% of what claude wrote back and type in “continue”
