Jordan | G2 Dynamics

Jordan (G2D) Founder • Quant Operator • Systems Builder AI-native R&D lab for trading + sports modeling Dallas, TX

[Jordan | G2 Dynamics @G2DJordan]

Building G2 Dynamics: A private lab integrating
AI, trading, and sports modeling under one system.

No hype.
No noise.
Only signal.

Early access to the Signal Room → DM “ACCESS”.

[AshutoshShrivastava @ai_for_success]

Disgusting, really.

What a cheap shot by Anthropic's Julian Schrittwieser at Jensen and Satya for supporting open source.

[Jordan | G2 Dynamics @G2DJordan]

I’m going to say this on every post about usage. It’s not the model it’s you.

[Eidzoku @evi77ain]

Codex might be seriously broken right now.

Yup, Codex rant part 3

GPT-5.6 Sol is burning through quota at a completely stupid rate and everyone notices it.

Here’s my usage graph:

Jun 23, GPT-5.5 xhigh: 327.8M tokens
Jul 2, GPT-5.5 xhigh: 549.9M
Jul 20, GPT-5.6 Sol High: 3.3B
I had a bunch of resets, sure, and several tasks were running. That let me burn more.
It doesn’t explain how 5.6 chewed through 3.3B tokens in one day. I genuinely don’t know what the fuck I would’ve had to do to get 5.5 anywhere near that.

And apparently it’s not just me.

@theo already covered 3 different ways GPT-5.6 was nuking usage. Looks like there might be a fourth.

5.6 uses the new Code Mode tool path. It can batch independent tool calls with Promise.all, but it barely ever does. Instead, it keeps running them one by one.

That means another model cycle for every call, dragging the giant context along with it every single time.

One trace found Promise.all in only 5 of 739 GPT-5.6 exec cells:

https://t.co/IQ6gqGWhSK

My own retained logs show the same shape. After removing fork-history replay, the 5.6 peak had 23,791 model requests. My clean 5.5 baseline had 3,803.

That’s 6.26x more requests, while tokens per request were actually slightly lower.

It wasn’t writing much bigger responses. It was just going back to the model over and over and over again.

Then someone tested explicit batching across two unrelated codebases. In the repeated High/XHigh runs, it cut model cycles by ~52–55% and weighted usage by 27–45%:

https://t.co/FGvWZqE4Xm

I’m not saying this one thing explains every token I burned. Several tasks were active, my local rollout logs aren’t the billing ledger, and we still don’t know how much of the spike came from agents themselves.

But both issues are still open, and this looks pretty fucking real.

For now, I’m testing explicit batching guidance. I’m also avoiding the built-in subagent/fork workflow for long-running work and moving bounded jobs into fresh disposable Codex tasks instead.

Self-contained prompt, fire and forget, inspect the result in the main task, archive it. Less context coupling, and each job is actually measurable.

I hope other people can validate it, too.

@thsottiaux @reach_vb @OpenAIDevs if it's true, it should be addressed. Cuz it might be THE cause.

[Jordan | G2 Dynamics @G2DJordan]

No 🤣

[Tibo @thsottiaux]

Should we rename ChatGPT Work to ChatGPT Vibe?

[Tibo @thsottiaux]

@shiri_shh Who is gonna tell him?

[Tibo @thsottiaux]

@Da7_Tech Oh no

[Jordan | G2 Dynamics @G2DJordan]

Tase is about perspective. You can’t please everybody. And there are work arounds. You are a frontend dev that doesn’t need a lot of reasoning just use the cheapest Claude you can afford and keep it moving.

Then the 2nd part. The limits are bad because you are lol your prompts leave to much room for interpretation. That’s leads the model to wander and go really, really deep. These newer models are different than last gen. They are more agentic so you can’t operate them like you did before, requires a more deft hand. Again just use 4.8 opus. These models are hard to “Vibe” with.

Start managing its usage and it should get much better and more narrow slices.

[Alex 💪 @awpthorp]

Noticed the same. On $100 plan, had Sol running for about an hour now and it's used 65% of the weekly limit.

Did have two resets from a few weeks ago so can't complain but definitely noticed a difference.

Basically rotating around 3 subs right now

$100 grok
$100 Claude
$100 OpenAI

[Jordan | G2 Dynamics @G2DJordan]

@fllowhiterabbit @pmddomingos Not very long at all when you have big giant factory size GPU farms.

[Jordan | G2 Dynamics @G2DJordan]

@mktpavlenko @kimmonismus It’s not real bro, it’s all propaganda lol

[Jordan | G2 Dynamics @G2DJordan]

@RijnHartman @Da7_Tech $100 plan every day heavy Ultra usage, burned one token. In 5 days

[Jordan | G2 Dynamics @G2DJordan]

One thing I love about Sol. Is it’s exposing what I’ve always known. Most of you are very bad at operating AI. If it is overthinking. You weren’t specific enough. Also other people have said it. You don’t have to run max intelligence. High and medium. Do 98% of what most of y’all are doing.

[Jordan | G2 Dynamics @G2DJordan]

Or less

[Tyler @rezoundous]

A GPT-5.6 Sol Ultra goal that runs for a day uses about 50% weekly limit, if anyone is wondering.

[Jordan | G2 Dynamics @G2DJordan]

I have no idea what you doing but I have been using Ultra heavy and I just burned my first token since they took away 5hr limit

[Johnny Jnr @jturntdev]

Cant just be who’s noticing this but GPT 5.6 sol in general drains your quota extremely fast.

How is it more efficient?

On 20x plan, you now get 1-2 days of usage. If you’re being conservative.

We’ve had alot of resets, this is just to stop everyone complaining.

When they all go, i see the beginning of the end.

They also reset limits 3hrs after we had to use our banked reset, due to expire.

[Jordan | G2 Dynamics @G2DJordan]

This is showing me what I have suspected for a very long time. All of you are horribly unorganized and have absolutely dog shit workflows. For the love of God Tibo please don’t listen to these people. We don’t need one tab for everything.

[Tibo @thsottiaux]

Now that we merged ChatGPT and Codex, what should we merge next? What's the double or nothing move.

[Jordan | G2 Dynamics @G2DJordan]

@AlexFinn @thsottiaux What you’re asking for is actually horribly unorganized. Also how are you? Consider considering yourself a technical person and you can’t even change tabs or chats?

[Jordan | G2 Dynamics @G2DJordan]

@scaling01 @thsottiaux 🤮🤮🤮

[Jordan | G2 Dynamics @G2DJordan]

@gazzzzzzman @thsottiaux 🤮🤮🤮🤮

[Jordan | G2 Dynamics @G2DJordan]

@_machi47 @BenPielstick @thsottiaux There’s nothing complicated about this the same chat button is there and it opens up a chat you just have other shit you can do. What are you even talking about?

[Jordan | G2 Dynamics @G2DJordan]

@metatransformr Xhigh…Not Max…not Ultra….XHigh…

[Jordan | G2 Dynamics @G2DJordan]

@ggg78g89 They can’t afford it, or else they wouldn’t have been taking it away in the first place

[Tibo @thsottiaux]

Morning. The last 48 hours of Codex and ChatGPT Work have been intense! Three important updates:

- Temporarily removing the 5 hour usage limit restriction for all Plus, Business and Pro plans
- Rolling out changes that will make GPT 5.6 Sol more efficient across the board and that will be reflected in less usage being used so that it can take you further. Exact impact to be quantified and shared
- We hit 6M active users, and are landing a usage reset in the next hour

Go do things