Darshan Amin

come for the best model, stay because we don’t treat you with contempt

@IndiaSpeaksPR yeah? what do you think about this then?

It is not easy for a man to rise from nothing to the highest corridors of power with a name like Lindsey. Credit where due.

hahahaha, long live @OpenAI

@claudeai saint @thsottiaux supremacy 🙏

We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.

@ChrisRMcGuire @TheWand68425335

NEW: The Department of Commerce just removed all export controls on AI chips on the UAE government and its AI national champion, G42. This is a HUGE deal, poses massive national security risks, will slow the US AI buildout, and will cause the largest data centers in the world to be built in the UAE instead of the US. These license requirements were in place because of significant national security concerns related to the UAE government and G42 in particular, which are well documented in public reporting. The UAE government and G42 can now buy as many AI chips as they want. Here's what will likely happen as a result of AI chips being uncapped to the UAE: (1) The UAE/G42 will buy millions of Blackwell chips (~6 million Blackwells are being made this year, so this would be a big chunk of global capacity); (2) UAE-owned G42 will build AI data centers globally, and be the first global competitor to U.S. hyperscalers, which currently dominate that market; (3) The UAE will build the largest data centers in the world, will become the world's second largest hub for AI compute, and leading AI models will be hosted in the UAE. This will pose enormous national security risks, as the UAE will become one of the most important AI compute hubs in the world, just as it now is for oil, but also a backdoor for China. It will create a massive foreign-owned competitor to the most important US companies. It will provide one of the few countries in the world with the money and interest to build AI models that rival the United States, with the tools to do it. And most importantly, it will exacerbate the already extreme supply shortage for AI chips in the United States, slowing the U.S. AI buildout. How does the United States benefit from this? How is this America First? This policy just seems like it's Steve Witkoff first, given G42 purchased 49% of World Liberty Financial in 2025 in an effort to convince the administration to let them buy large numbers of AI chips. That seems to have worked. It is hard to overstate how far-reaching the implications of this decision could be for the global AI infrastructure buildout, which was previously dominated by the US--but now has legitimate non-US competitors that have extensive ties to China. And the AI chip supply crunch for US firms will get tighter, making it even harder to build data centers in the US. Why are we doing this? Who benefits?

@thsottiaux @ClaudeDevs saint tibo 🙏

@ClaudeDevs I smell fear

We've reset 5-hour and weekly rate limits for all users.

@nickporwal @gauravh1 @TANCHJIM_CN very cool, thank you so much.

@darshansamin @gauravh1 @TANCHJIM_CN You can check audiostore India website. I have ordered many times from it. Its reliable.

Tanchjim Bunny DSP! 😇 Thanks to @nickporwal for a fantastic IEM suggestion! 🙏🏻

@nickporwal @gauravh1 @TANCHJIM_CN hey, thanks for this - will order. i heard a lot about tangzu - wan'er sg 2 red lion too, any views on that? also, if you could point me to a good place to order from that'd be great. thanks.

@gauravh1 @darshansamin @TANCHJIM_CN Buy bunny. Chu 2 is also very good. No doubt. But bunny is better in my opinion. It feels more open & clean. Also moondrop is planning for chu 3, you might want to wait for that also.

@gauravh1 @nickporwal @TANCHJIM_CN thanks, will check

@darshansamin @nickporwal @TANCHJIM_CN Check this review. youtu.be/27pTVevAlK8?si…

@nitish_fixaeo looks sleek 🙌

pretty sure AI doesn't even know your brand exists. type your domain. 10 seconds, free. prove me wrong 👀

@CrankyFed @TheWand68425335

Your pub's flag isn't on the moon

@midasavocado @thsottiaux 🤣

We need to be able to choose what this number is. Doing this will allow users to more effectively utilize Codex in their day-to-day lives. There would be a button where we can just add 1 to this number under Settings -> General Thank you for your consideration and I look forward to seeing this in the next update.

This is a good feature request. We will take more feature requests like this one. The combination of {motivation, explanation, screenshot} forms a perfect prompt for Codex. Thank you.

@thsottiaux Thank you Saint Tibo 🫶🤎

Codex usage limits will be fully reset again in the next hour and we will credit one additional reset into your bank for your own usage over the next 24 hours. We investigated reports that Codex usage was being consumed faster than expected. There wasn't one central issue, but a few smaller problems compounded for some users. Here's what we found and changed: - Actual usage: Auto-review had become more proactive, another change was triggering more subagent work, and background suggestions could run twice or retry too frequently after failures. We reverted the changes and fixed suggestion scheduling, duplicate generation, and retry behavior. This should reduce unnecessary background token consumption while preserving the work users explicitly request. - Usage reporting: Auto-review was incorrectly appearing as GPT‑5.4 usage, and failed or rate-limited requests were still shown as turns. Auto-review now appears as its own category, and only successful requests count toward the turn graphs. Rate-limited requests were never charged, but they were being displayed incorrectly. - Immediate relief: We reset usage limits while rolling out the fixes, then shipped hotfixes across the CLI, desktop app, and usage backend. - What to expect: New usage data should be clearer and actual consumption should be lower. Historical charts may still show auto-review under GPT‑5.4 because older turn data was not relabeled. Features that intentionally perform more work; such as /goal, subagents, and higher reasoning levels will still naturally use more capacity. All fixes are now deployed, and we've added more detailed monitoring so we can detect background-usage regressions sooner. We'll continue watching the results closely. Thank you for building and doing all sorts of things with Codex.

@nitish_fixaeo 🤣

@darshansamin not just that, InsiteChat.ai as well.

the one technical decision you'd take back. every senior engineer has one. what's yours?

@ajambrosino 👀

what happened in Feb? 👀

@nitish_fixaeo 🤣

saas milestone: 50 signups 🎉 (38 are me in incognito, 9 are bots, 3 are real humans who bounced in 4 seconds. but still. 50.)