Joined May 2022
Hey @thsottiaux, Codex feature idea: let users roll over 50% of unused weekly limits. I often burn through my remaining usage before reset just to avoid wasting it. A capped rollover could smooth compute demand, reduce pre-reset spikes, and improve UX. Win-win?
67
Replying to @jxnlco
7
17
7
307
58,605
I really don’t know how I do it
1
52
the insane part is I feel gpt 5.5 is smarter than me in some subjects but purely from vibes and pattern recognition I’m able to navigate those subjects with it and paint the context window in such a way that only the most delicate tokens come out
1
50
I don’t understand why I feel so good at vibe coding compared to my peers. it just works with my brain
38
I’m either completely retarded or genius, there’s no in between
25
I’m seriously so incredibly cracked
1
18
Replying to @thsottiaux
@thsottiaux is there any documentation or research that shows the performance differences of codex if you keep switching thinking modes? Or use codex x-high on task that really should be for low?
1
38
is there any best practice that the team has noticed?
19
very interesting how the people I follow are either some ai maxxing adopters or quant doomers
36
LuckyNonce retweeted
fireworks in celebration of re opening
LOUD EXPLOSIONS HEARD AROUND THE STRAIT OF HORMUZ — FARS NEWS
9
62
1
1,005
70,540
lol I knew it was bad but not this bad
JUST IN: 🇺🇸 President Trump traded stocks over 3,700 times in Q1 2026 - averaging 59 trades per day, 9 per hour, or one trade every 7 minutes.
46
let’s not forget
$INTC has a lot of room here...its like 15 beers behind and just got to the party.
1
11
4,407
back months in CL are looking more and more bullish. Market finally realizing this won’t be resolved in a headline?
1
48
Replying to @SIGKITTEN
@SIGKITTEN do you feel that codex 5.5 xhigh does better once the codebase is 70% there and it needs to finish/clean up? Or does it do better starting from a spec?
1
37
current workflow: > chat with low about requirements > give thread to xhigh to create spec > breakdown spec in smaller task per worktree > run medium on each worktree in parallel with task > merge worktrees with xhigh
23
he just like me for real
give me six hours to do a task and i will spend the first four hours refining my agentic coding workflow
2
43
LuckyNonce retweeted
gas is $5 i am not coming to just cuddle
157
4,322
415
53,274
1,352,375
I bought a few europe trips today seemed like relative value. if they run out of jet I triple money with insurance else I get vacation
21
when the hooker asks you how you're gonna pay
159
283
49
8,107
448,420
wealth manager told me last month to sell energy and buy small caps btw
23