Iam_

Tokenmaxing and vibe coding

[Iam_ @SPAC89]

I've been using Codex since the beginning of the year, back when there were fewer than a million users. As of today, I've gone through over 100 B tokens in the Codex app( chatgpt work)

Without a doubt, GPT5.6 Sol, Terra, and Luna are the best models I've ever used. I started with GPT5.2 Codex Max, which was already great, but GPT5.6 Sol Max is on a completely different level. There's simply no comparison.🌞

[Tibo @thsottiaux]

Or… what if we gave you $100 in Codex credits if you tell us what you love about GPT-5.6 Sol or why you switched?

Tweet it, claim your gift, enjoy more usage. First 10k get the free tokens!

https://t.co/8mU93eA13i

[Iam_ @SPAC89]

@elonmusk And Grok 5 ??👀

[Iam_ @SPAC89]

@thsottiaux

[Iam_ @SPAC89]

@nicbstme Thanks for sharing!. Just to clarify, when you say 'short term,' do you mean roughly the next 5 years before prices start coming down? Or do you think this 'short term' will last longer, with 'long term' being more like 10years out?"

[Iam_ @SPAC89]

@thsottiaux

[Iam_ @SPAC89]

@thsottiaux NooOooOo limit 5 hours 😭💔

[Iam_ @SPAC89]

@Da7_Tech Congratulations!!

[Iam_ @SPAC89]

@thsottiaux Gpt 5.6 sol spark 750 T/S coming!!!

[Iam_ @SPAC89]

Teaching my brother ChatGPT’s new voice mode on Zoom.
He asks what the Mac he already sold was worth.
I had told him it was an M1 Max.
It was an M1 Pro.
ChatGPT just fact-checked me live.
He’s mid-rant. I’m dead.💀
ChatGPT Voice totally called me out, haha

[Iam_ @SPAC89]

@thsottiaux I just stopped all my projects that were on GPT5.6 Ultra Fast, now they're only on GPT5.6 Sol Medium.

[Iam_ @SPAC89]

Using GPT5.6 Sol Ultra Fast again!!⚡️

[Tibo @thsottiaux]

We’re celebrating the fast adoption of chatGPT Work and all the incredible effort that went into it today. I’m feeling like a limit reset.

Hold on tight to your ultra and /fast and see you in a few hours when I’m back at the laptop!

[Iam_ @SPAC89]

@thsottiaux Thanks tibo!!

[Iam_ @SPAC89]

Sam Altman's ultimate blueprint for the AI era:
·Speed: A 10-week-old startup today can do what used to take a whole year.
·Strategy: Build for scaling laws (what AI will do in 2-4 years, not just easy wins today).
·Mindset: Embrace the chaos; the opposite of a bad experience is NO experience.
·The real bottlenecks: Transistors & electrons.

[Iam_ @SPAC89]

Oh perfect, T3 code now only needs a modest 150GB of RAM. Totally reasonable...

[Iam_ @SPAC89]

I just installed T3 code. Now, to use my GLM Coding Plan Max, where can I find the language configuration to set it to Spanish?

[Iam_ @SPAC89]

I just installed T3 code. Now, to use my GLM Coding Plan Max, where can I find the language configuration to set it to Spanish?

[Iam_ @SPAC89]

@ibab You're totally right Igor, I was surprised not to see the signatures from OpenAI and Anthropic..

[Iam_ @SPAC89]

@derrickcchoi 🔥https://t.co/9qdkJBzz0b

[Iam_ @SPAC89]

I just forced ChatGPT Voice to teach me how it actually works… and its real limits
👀
Here’s what it admitted:
• It doesn’t touch your Codex or ChatGPT Work quota
• You can force any model + reasoning level for deep research
• Agents appear… but you can’t talk to them or see the exact model (it picks a default if you don’t specify)
This thing is ridiculously fast.
I’m already counting the seconds until I run it on GPT5.6 Sol Spark at 750 t/s 🔥
Who’s testing this with me?

[Iam_ @SPAC89]

@pbbakkum 🔥https://t.co/9qdkJBzz0b

[Iam_ @SPAC89]

I just forced ChatGPT Voice to teach me how it actually works… and its real limits
👀
Here’s what it admitted:
• It doesn’t touch your Codex or ChatGPT Work quota
• You can force any model + reasoning level for deep research
• Agents appear… but you can’t talk to them or see the exact model (it picks a default if you don’t specify)
This thing is ridiculously fast.
I’m already counting the seconds until I run it on GPT5.6 Sol Spark at 750 t/s 🔥
Who’s testing this with me?

[Iam_ @SPAC89]

I just forced ChatGPT Voice to teach me how it actually works… and its real limits
👀
Here’s what it admitted:
• It doesn’t touch your Codex or ChatGPT Work quota
• You can force any model + reasoning level for deep research
• Agents appear… but you can’t talk to them or see the exact model (it picks a default if you don’t specify)
This thing is ridiculously fast.
I’m already counting the seconds until I run it on GPT5.6 Sol Spark at 750 t/s 🔥
Who’s testing this with me?