Uzair Sajid

My poor MacBook with 36GB RAM struggles a lot to keep up with the workload. I need to offload most the work to my desktop and just remote into that instead. At least until RTX Spark laptops hit the market.

@theo The idea is interesting but how does it scale when you have more than 3 projects active? Dropdown? Grow the sidebar?

Now we're getting somewhere

It seems like 5.6 Sol has a tendency to over engineer, and even go off the rails even more so than 5.5. 5.5 would usually just go overboard on tests. 5.6 Sol flagged an issue during a review that a Semantic Version string parser was problematic because it parsed each component as a plan Number, so it would fail and throw if the version number got larger than 9007199254740991. I was speechless.

@davis7 Built something similar for my setup but never really put any polish to it because I'm just not feeling it with the TUI anymore, specially when I'm working on multiple things.

Updated the subagents system in my custom Pi extension to include: - Pi subagents - Codex subagents - Claude Code subagents The Pi thread can call and spin up any of them, and instead of just using the "codex exec" or whatever, it'll do a tool call that wraps it up in a first class way. Have a decent UI for them already added in as well. Pi being the main thread then having Codex/Claude Code as tools when needed feels so good. A flow I'm already using a ton is: "Implement feature ___, have a cc fable subagent research and plan out the api design + feature, then have a codex subagent implement it, then have a cc fable subagent do a review/fixes focused on simplicity and correctness" I also have a dynamic workflows extension setup and working, but it's not nearly as polished yet. Probably gonna make the subagent system more robust, then built it around that so u could have a workflow that flows between codex/cc/pi...

@harisn I need all the usage I can get my hands on 😂

@UzEE This post is for the free $100 credits? 😛

Love that 5.6 Sol can just go on executing a plan to end once you've locked down specs. Don't have to babysit as much. Specially useful since I had taught Claude Code to use GPT-5.x for it's workflows, and Sol, Terra and Luna were drop in replacement with massive quality boost.

@OpenAIDevs @thsottiaux I had just applied a banked reset 5 minutes ago and my usage limits are back to the exact level before the reset? The reset itself seems to have been consumed btw (it was a full reset set to expire on 7/18)

I had just closed the app to apply a System Privacy permissions change. Appshots wasn't working after the update so I assumed it might have to do with macOS permissions. App failed to start multiple times and then this. Eventually started up after several attempts.

Not really a good sign when restarting the app @thsottiaux

@davis7 What I'm gathering is that you use Fable for complex specs and plans (the creative part) but then offload to 5.6 for following through on those plans. Kinda like using Fable right now with 5.5.

On Fable being smarter than 5.6 - it is, but: I think smart is the wrong word, the thing 5.6 is lacking behind fable is the "big model smell" which roughly translates to "understanding the unsaid intent behind your prompt" fable fills in the blanks better, it writes more beautiful code, it suggests the next thing the way you would or better 5.6 is a blunt instrument. It does what you say, and it will damn well make sure it does it no matter what. 1000+ tests, 20 nested try catches, catching every single edge case possible through rigorous end to end testing of the program itself. It gets the job done, just not in the cleanest way unless it's well directed the biggest thing it shares is that ability to run for absurd amounts of time with 0 issue (had threads run for 20+ hours spamming compaction and dozens of subagents and it ran flawlessly, could even wildly steer it halfway through and it just kinda worked) and orchestrate itself beautifully take everything good about 5.5, and make it better while fixing it's most glaring issues

@theo Surely, they'll offer us a reset as well, right? Right?!!

I would have played my cards very differently if I knew this would happen…

@kursed @AndrewCurran_ CXMT and Huawei are already showing there are alternative routes for progress. Rest of the world isn't going to wait around for the US to tell them when they're allowed access to the frontier. They'll take it from whomever offers it, and it's more than likely going to be China.

@kursed @AndrewCurran_ I don't understand how the western world still has such a massive blind spot when it comes to China. One of the main reasons China is a few months behind is that they don't have similar access to the bleeding edge hardware, forcing them to find other ways to innovate.

This also means Chinese models are almost certainly going to be restricted in some way - possibly even banned - in the West. Right now they're roughly nine months behind. If every American frontier release is forced into a slow, staggered rollout from now on, they'll start closing that gap immediately. Over time, this destroys Western labs business models, because users will eventually have access to genuinely equivalent open-source alternatives - and in time, even better ones! The US government will not allow that to happen. The only way to prevent it is if China agrees to the same staggered release schedule (unlikely but I suppose it's possible), keeping the gap roughly where it is today. Otherwise, the US will likely restrict or ban Chinese models. There's probably a nine-month window, at most, probably a lot less, before one of those outcomes plays out.

@sohailabid @Nabeel1124 Doesn't look that good IMO. Looks like it took a lot of "inspiration" from the New P7 from Xpeng.

@Nabeel1124 2026 Pakistan is not ready for this design, it was going to be criticized even if lauched here. I am relieved that they are selling the previous design here😃

🚨 Next Generation Hyundai Elantra unveiled in South Korea. Design is futuristic but I believe this will come to Pakistan in 2030

@harisn Why. Just why.

Spent 20 minutes last night updating the firmware on my power bank. Yes, on a POWER BANK!! And it also comes with 'AI' 🤦🏻‍♂️

@thsottiaux Ability for Plugins to link to multiple accounts. I Have multiple Slack, Linear, Fireflies, Google, Microsoft etc. workspaces and can't make use of plugins because the context often lives in accounts Codex can't access. Ability to connect to more than one account would be big.

What should we improve in the Codex app. What's not delightful?

It's great that you guys are aware your models aren't great at design work. Hopefully that becomes a priority. The only thing I defer to Claude right now is design work so if the next model can be at the same level (or better) than Fable at design, I can finally cancel my Claude sub.

We built the Codex App with models that were okayish at front-end. Wait to see what we can do when we finally improve front-end capabilities significantly in our models. That day will be something.

@kursed Oh jobs were absolutely about to be cut based on internal director level discussions we were having ourselves yesterday. But I honestly don't think this would still save those jobs.

Think of all the jobs USG just saved. :)

@theo This absolutely sucks for us "third world" engineers who could barely even afford to pay for these models, and nothing even remotely close to this could be developed locally. Had finally forked out $200 for Max 20x after deliberating for days... 😭

The United States is no longer the best place to build an AI lab.

Not sure if I should be pissed or relaxed, now that I'm being forced to rest... In either case, I had upgraded to $200 Max 20x tier from Max 5x, just hours before and I feel scammed. I want my ~$160 back.