Already loving the new Intelligent UI feature in ChatGPT 😍
With Intelligent UI, GPT‑6 can now compose responses using text, visuals, and interactive elements, choosing how they fit together based on your question.
Responses can include graphics and charts to help explain an idea, along with tappable buttons, forms, and interactive experiences you can use directly in your conversation.
1
160
waaaaaaaaaaaait a minute... GPT... *6* Sol in ChatGPT? instead of... 6.*1* Sol? 😔
GPT‑6 with Intelligent UI rolls out globally to Plus, Pro, Business, and Enterprise users today and will expand to Free and Go users starting tomorrow.
GPT‑6 in ChatGPT is powered by GPT‑6 Sol for Plus, Pro, Business, and Enterprise tiers, and GPT‑6 Luna for Free and Go tiers. Both are tuned for everyday conversation.
This update applies to the Chat tab in ChatGPT. The models powering Work and Codex are not changing as part of this release.
openai.com/index/gpt-6-for-e…
1
152
Years ago, some people were trying to get iOS running on the HTC HD2
They worked on porting XNU, but didn't finish
Now, they have iOS 7 running on the HTC HD2 (assisted by Codex) 👀
I can't recall ever seeing real iOS booted bare metal with working springboard on a non iPhone
8
1,589
Until today, there was an argument to be made in favor of not using "Approve for me"/auto-review in Codex simply because it consumed your ChatGPT Work/Codex 5-hour & weekly usage quota
now, it's finally free, just like Claude Code's equivalent feature has been for a while! 🥳
Day 2.1/
We have made Auto-review free for all users signed in through a ChatGPT account. You can enable it in settings > permissions > auto-review. Auto-review improves upon the default sandbox setting that requires you to approve everything, which is prone to decision fatigue unless you spend a lot of time configuring specific rules.
It allows you to run long tasks while having a second agent review all actions taken by the primary agent. Its only goal is to prevent high-risk actions from being taken and to protect against unwanted actions that are not aligned with the original user intent. This Auto-review feature is now free and does not draw usage from your plan.
1
832
🚨 BREAKING: Microsoft confirms on publicly accessible web page that OpenAI has been using Looped Transformers in the GPT-6 series, proving The Information's reporting was correct all along‼️
GPT-6.1 Sol uses 2 inference passes, with a passing mention of "instead of three" 👀
78
196
3,461
824,642
For those confused by "same base model weights as GPT-6 Sol", I think MS meant 6 & 6.1 are both post-trained models on top of the same pre-trained "base model", not that the final weights are identical
So different post-training (+ one less loop)
🚨 BREAKING: Microsoft confirms on publicly accessible web page that OpenAI has been using Looped Transformers in the GPT-6 series, proving The Information's reporting was correct all along‼️
GPT-6.1 Sol uses 2 inference passes, with a passing mention of "instead of three" 👀
7
3
85
68,234
I swear, every time I try using an LLM with the recommended reasoning effort setting for literally any use case I'd use an LLM for, it always ends in pain and misery... as soon as I crank it up to Max™, it's fine
this applies even with Opus 5.5 (the best LLM I've tried by far)
2
356
Damn, first OpenAI, now Google's running out of compute too? How on earth does Anthropic (formerly compute bankrupt) seemingly have more compute than Google & OpenAI now? 🤔
117
it has been 4 hours since the time OpenAI was supposed to reset everyone's ChatGPT Work/Codex usage quotas
no reset, and the announcement got a community note as a result
I wonder if OpenAI is trying to fix the super low TPS for GPT-6.1 Sol first? it's slow even on fast mode 🤔
Global reset landing tomorrow 10am PST for all paid ChatGPT accounts. Apologies for the slow start with GPT-6.1 Sol, it's now back to running at expected speeds after the massive load spike in the first two days.
3
2
484
I've come across *many* posts so far from Googlers implying Gemini 4 Argon isn't benchmaxxed, & here, someone says it outright
it appears Bloomberg's reporting *might've* been wrong 👀 🤞
Honest Argon take: it's not perfect (Opus 5.5 thinking traces are prettier), but it's the first big model we've released that's been this battle tested.
Far from being benchmaxxed, the 200k+ Googlers who rely on it everyday ensured that real utility was prioritized.
1
120
Gemini 4 Argon's benchmark results looks fantastic
here's hoping that translates into real world use this time... (and that it actually ships unlike Gemini 3.5 Pro 🥲)
apparently, Google is using it right now to migrate C/C++ codebases to Rust, including Fuchsia OS's kernel
Introducing Gemini 4 Argon – our new frontier model.
It’s built for complex workflows across coding, enterprise knowledge work, and cybersecurity defense – rolling out today to a set of trusted testers through our Fairwind Program.
73
so GPT-6.1 Sol is Astra Minor lol
meaning... we can (presumably) now see how much sequence-level recurrence helps compared to no recurrence at scale!
helps token efficiency a *ton*: openai.com/index/introducing…
GPT-6.1-Sol has insane CoT-controllability compared to 6-Sol and 5.6 Sol
this pretty much confirms it that GPT-6.1-Sol is their smaller looping model
1
324
anthropic is so compute prosperous now apparently that I got an email offering me 50% off my first month of Claude Pro as a returning subscriber 👀
215
If we assume Astra's arch is like the Full-bandwidth transformer or Recurrent Looped Transformer, and that the model is larger than Sol,
could Astra-*minor* be the size of Sol, but with recurrence? if so, the benchmarks could show how much sequence-level recurrence in a Transformer buys 👀
I decided against writing an article and revealing how I come to these new estimates. I want to do more analysis on the coming models.
my latest central estimate is that GPT-6 Astra is:
- 4.2T@120B
- around 112 layers
- loops 50% of its layers once
I also have evidence that GPT-6 Sol and Luna, and Opus 5.5 are NOT looped language models.
The leaks about "Astra Minor" suggest a smaller Sol sized looped model coming, that probably slides right into the $20-25 price point.
167