Ruby @rubywwwilde

Building a tool for thought. Joined January 2026

Tweets

452
Followers

26
Following

111
Likes

2K

Ruby @rubywwwilde

3 days ago

@badlogicgames do you know any other social media that are better than twitter? it's so weird how we know that platforms we use fail to deliver what we expect from them, and still, here we continue using them.

0 0 0 425 0

View Details

Ruby @rubywwwilde

4 days ago

@bundeskanzler @POTUS cringe.

0 0 0 2 0

View Details

Ruby @rubywwwilde

5 days ago

@dbreunig Huh. Very interesting, I have something similar in my system prompt, but not about copyright of course lol. Self-check thing does work but having another agent in the loop helps much better—though this increases time before final response a lot.

0 0 0 94 0

View Details

Ruby @rubywwwilde

5 days ago

@Mihonarium I mean. It's braindead. It bases answer on a link that states the opposite.

0 0 4 565 0

View Details

Ruby @rubywwwilde

a week ago

@ValerioCapraro skill issue. it's quite sad, because I agree with the underlying take, but this example is so bad.

0 0 0 291 0

View Details

Ruby @rubywwwilde

2 weeks ago

@iydheko @uwukko translation is wrong by the way. it says: dick dick hey ai but there was no "hey" in the original tweet.

1 0 3 83 0

View Details

dax @thdxr

2 weeks ago

how to be good at your job - realize this one thing is actually made up of two separate things - realize instead of solving the direct problem you can solve a broader problem - instead of implementing thing, implement other thing that makes it easier to implement thing

130 240 4K 124K 1K

View Details

Ruby @rubywwwilde

2 weeks ago

@thdxr wow. you seem to be accidentally describing how to be good at meditation.

0 0 0 418 0

View Details

Ruby @rubywwwilde

3 weeks ago

@eoinyk Lol, the community note did not help with this one.

0 0 0 3 0

View Details

Ruby @rubywwwilde

3 weeks ago

@theo So the skills in Hermes are not "enable what you use" but "disable what you don't use"? Appears to be the opposite of pi-mono in this matter.

0 0 1 318 0

View Details

Ruby @rubywwwilde

3 weeks ago

@deepfates wtf. mine gave me "Cesare, from Pavese's The Moon and the Bonfires.", when I asked who's that, it told me "Sorry, actually Cesare is the author of the book" > *looks into Claude's thinking* Not a single mention neither of this book, nor of the author.

0 0 2 693 0

View Details

Ruby @rubywwwilde

3 weeks ago

@uwukko what about imputnet/hueblyadskaya-pizdoproebina?

0 1 3 156 1

View Details

Mitchell Hashimoto @mitchellh

3 weeks ago

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem. As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)! I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work. It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results? 88ms => 1.5ms 150K allocs => ~500 allocs Incredible right? Nope. My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path. This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput. The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity. Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

308 978 9K 794K 2K

View Details

Ruby @rubywwwilde

3 weeks ago

@0xRogerlee Looks really nice.

0 0 0 3 0

View Details

Ruby @rubywwwilde

3 weeks ago

@tjthedev @dvdeijk @stupidtechtakes Someone should build MythOS then!

0 0 0 10 0

View Details

Ruby @rubywwwilde

3 weeks ago

@elder_plinius Yes, the same tactics work again and again, it's just sometimes you need to craft compelling enough setting to make the request actually valid. And it's really hard to protect against such jailbreaks, like how would you even do it without lobotomizing the model.

0 0 5 5K 2

View Details

Ruby @rubywwwilde

3 weeks ago

@claudeai Is it the same tokenizer or no?

0 0 0 5 0

View Details

Ruby @rubywwwilde

3 weeks ago

Wow. I've opened @opencode for the first time and it looks gorgeous.

0 0 0 13 0

View Details

Ruby @rubywwwilde

3 weeks ago

@uncreativetom Hmm, I am not convinced by the drop resistance scores. I am carrying my iPad pro 13, without a case and it had a lot of crazy falls over the last two years. It has very little scratches.