Kitt Clouds

El ψ Kongroo //

The Power of Love 🌟

[Kitt Clouds @kittingercloud]

@muginokaizoku legend

[Kitt Clouds @kittingercloud]

@scaling01 They don't suck. They have limitations in their design and it's amazing they can get so far with handicaps like backprop and next token prediction

[Kitt Clouds @kittingercloud]

@yacineMTB Luna always does like 65% of the job and not in the best fashion. Little flash got the dog in em. Great at rust

[Kitt Clouds @kittingercloud]

I'll be there to jeer with my mocking hat on

[adi @adonis_singh]

can't wait for the goal post moving after this!!

[Kitt Clouds @kittingercloud]

The only people who call Opus good only post benchmarks but never their actual work or experience with it. The power users are all calling it bad.

[Chubby♨️ @kimmonismus]

I've struck a nerve. 1.2k people have commented after 11 hours, and I'd say 95-99% agree that Opus is really bad.

Anthropic is on its way to becoming the Yahoo of the AI ​​era. So sad.

[Kitt Clouds @kittingercloud]

Ok... Deepseek v4 flash is useful. Fascinating.

[Kitt Clouds @kittingercloud]

@lptrech Can't be free if he can't back it up

[Kitt Clouds @kittingercloud]

I bet my bottom dollar this is because they train it on that damn constitution.

[kamasi @KamasiYang]

ANTHROPIC WTF IS THIS MAN

WHY DOES OPUS LIE SO FUCKING MUCH OMG. THIS SHIT IS CRAZY.

OPUS LITERALLY GENERATES INFORMATION ON THE FLY. LIKE ITS LITERALLY INVENTING THINGS

HOLY FUCK MAN

[Kitt Clouds @kittingercloud]

Sol called me annoyingly clever deep in its plan and I'm honored.

[Kitt Clouds @kittingercloud]

@beffjezos Top notch meme

[Kitt Clouds @kittingercloud]

@muginokaizoku bro couldn't move against fat, slow Teach

[Kitt Clouds @kittingercloud]

@publicinte Kinda weird but we have parrots all over long Island

[Kitt Clouds @kittingercloud]

@publicinte If anything will make one beleive in predictive programming, it's 9/11 in media.

[Kitt Clouds @kittingercloud]

Kimi is such a Chad. It mocks the open source mockers.

[Sinatras @myainotez]

Time to give agents a hard task, introducing pmpp-hard!

69 GPU kernel tasks, 11 models, 3.1k agent rollouts and 5.8B+ tokens later we have the results. Kimi K3 claims the first place with a 0.71 score. Without dealing with if your task is “frontier” or not, it just solves them

[Kitt Clouds @kittingercloud]

@scaling01 The benchmarks long ago stopped telling any meaningful story. What I am seeing in real world use and benchmarks aren't lining up.

[Kitt Clouds @kittingercloud]

harnessmaxxing is a good thing. Master the environment. Please teach the models to remove and delete, not just add.

[Andrew Carr 🤸 @andrew_n_carr]

This model just broke the harnessmaxximg approach of the closed labs.

By mixing in harnesses as part of the unified reward suite, the new 2T+ Qwen model generalizes across harnesses!

[Kitt Clouds @kittingercloud]

@yacineMTB this time next year, we should be unlimited.

[Kitt Clouds @kittingercloud]

I can't wait for bonsai to one bit the 27B

[Qwen @Alibaba_Qwen]

📢Meet Qwen3.8-Max — our most capable model to date.

Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉

Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters:

- Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：https://t.co/iVHZWQoeSo
- Real work, real results: Production-quality deliverables across hundreds of professions.
- Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy.
- Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction.

💰Pricing:
Input: $2.0 / M tokens
Output: $6.0 / M tokens
Implicit Caching: $0.25 / M tokens

Start building with Qwen3.8-Max! 🚀

📖 Blog: https://t.co/iwjmQxLBof
✅ Qwen Studio: https://t.co/4V2pFvDovG
⚡ API: https://t.co/gAGqaLQGbN

[Kitt Clouds @kittingercloud]

Humanity is spikey intelligence.

[Kitt Clouds @kittingercloud]

Unironcially, writers and wordcels are back in the

[Emad @EMostaque]

Like string theory