ctxc

4 ms·

1.
Tell HN: Claude Code appends new commit signature even if older settings is off
8 points by ctxc 2mo ago|2 comments
2.
by ctxc 3mo ago
Thank you for your comment! It was useful. Found this that helped me look through the signs https://www.reddit.com/r/SDAM/comments/nkolwi/my_attempt_at_...
3.
by ctxc 3mo ago
It's insane, right? Reverse engineering is so much fun, I find myself doing it on random apps so often now
4.
by ctxc 3mo ago
agreed with "the final step". I worry adding commands like these might affect quality of responses at each step, which then stack up to result in an overall worse outcome.
5.
by ctxc 3mo ago
Probably because they skew towards higher information density
6.
by ctxc 3mo ago
"LemonLime bills itself as a no-code AI startup, not to be confused with the hundreds of startups building something similar." Ouch
7.
by ctxc 3mo ago
20 games, of 104? Not to mention games were played in indoor air conditioned stadiums. MLS already has a system where a game has a break only when it reaches 28deg C for a period of x minutes
8.
by ctxc 4mo ago
I didn't think about this parallel, but that's just so sad.
9.
by ctxc 4mo ago
Thanks - I didn't verify this personally, but still stand corrected. If training wasn't considered outside the law, this goes on to make the point about double standards for US vs Chinese model training methods.
10.
by ctxc 4mo ago
Anthropic was asked to pay book authors because it trained on pirated downloaded books. What about books and art where the author/artist does not authorise AI to train on it? They do happily train on it, ignoring their "ToS"
11.
by ctxc 4mo ago
Nice work!
12.
by ctxc 4mo ago
It was always this way!
13.
by ctxc 4mo ago
Yeah weird, bc I just checked and it isn't oss. Only the SDKs are OSS, which...doesn't help much
14.
by ctxc 4mo ago
/s not needed, it's the easiest path
15.
by ctxc 5mo ago
(you mean it's _not_ AI generated)
16.
by ctxc 5mo ago
I feel like with time, 50 is the new 70.
17.
by ctxc 6mo ago
Those numbers make sense to me...120 million input tokens is like 120 sessions of hitting the full context limit, which seems like a lot to me though
18.
by ctxc 6mo ago
Haha! Made me chuckle
19.
by ctxc 6mo ago
Reads like a LOT of tokens to me. What does your usage /workflow look like? I'm v curious because although I do use Claude code, my token counts aren't nearly as much I want to know if I'm missing something cool!
20.
by ctxc 6mo ago
When you have a general idea of what smells bad vs what's okay...why? I'd rather get it from the LLM and review
21.
by ctxc 6mo ago
They report an hour less than average sleep time the first 3 months?! How did get so lucky?
22.
by ctxc 6mo ago
WOW YES OMG do you remember what the wrestling kinda game with ed ed and eddy was?
23.
by ctxc 7mo ago
Out of curiosity, what kind of custom harness are you thinking?
24.
by ctxc 7mo ago
I do something tangential. If you can pick out identifiers for these pages with deterministic features (financial pages have the most numbers on the page, or have the word "director", "general manager" and "managing
25.
by ctxc 7mo ago
Scale, my friend, scale...
26.
by ctxc 7mo ago
Interesting, I'd love to know more. Are parts of it public?
27.
by ctxc 7mo ago
haha, sweet. One of the cooler things I've read lately
28.
by ctxc 7mo ago
I was going to type you a sympathetic message and took a peek at your profile, and...well that's quite a resume you got there!
29.
by ctxc 8mo ago
Accountability. Same reason I want to read human written content rather than obvious AI: both can be equally shit, but at least with humans there's a high probability of the aspirational quality of wanting to be considered "good&q
30.
by ctxc 8mo ago
I took ~ to be a "singing tone" for some reason till I saw sibling and realized it might be an attempted strikethrough xD

More ›