Muhammad Yusuf Rakhimov

Everyone wants “AI that codes for them.” Nobody wants to do the boring part: define context. That’s why most AI-generated code breaks in production. Here’s the real explanation 👇🧵

@robinebers Definetely gpt 5.5

what AI model designed this? Opus 4.7 of GPT 5.5? 🫠

@brillaas Meta

I just filmed a 50-minute Meta Ads training for 2026. (full course) This has everything I've ever learned after - Making $50K in the first 20 days on a BRAND NEW ad account - Working with 40+ businesses - Generating over $20M+ through our funnels And it’s going to help you overcome the 3 BIGGEST mistakes people are making with their Meta ads in 2026. 1. Generic ad messaging 2. Incongruent funnels 3. Poor pre-call systems Thing is... They're EASY to fix when you know the system. And I compiled quite literally everything I know into this training. Inside: 1. Complete Andromeda creative strategy (how to build 25-30 diverse ads) 2. Funnel structure and how to setup your landing page for paid ads 3. Pixel conditioning done right to get qualified leads only 4. Pre-call nurture system for 70-80%+ show rates 5. Campaign structures for EVERY budget 50 Minutes of pure sauce. If you want it: Comment "META" And I'll send the full course. (must be following for DM)

@abouelatta_ali not working with codex

I hated how bad agents are at design I hated how Codex can't access Mobbin So I created Lazyweb - 257k+ screens (apps/web) - 6 opinionated design research skills - 1 MCP (Claude/Codex) 100% Free...AI native...no rate limits...no subscriptions.. Enjoy (and tell a friend) 🫡

@pmitu For almost 2 years, i am paying for 3 domains just not to lose them😃 but build something myself

Builders How many domains do you own right now?

GPT-5.5 is on par with Claude Mythos - GPT-5.5 average pass rate of 71.4% (±8.0%) - Mythos Preview 68.6% (±8.7%) - GPT-5.5 solved a task that takes a human expert ~12 hours in under 11 minutes at a cost of $1.73

@sama can't wait to try

big upgrade for codex today! try it for non-coding computer work.

@muxlisa_ms @ahroralikhanov Eng zuri grafika uchun gpt images 2.0

@muxlisa_ms $100li tarifi yetmidi. Ayniqsa opus 4.7 juda token hungry model. Codex $100 undan kura ancha yaxshi

卧槽，GPT 有点强的离谱 昨晚看到@ianneo_ai 的PSD 导出，尝试去试一下，发现不止 PSD 文件能导出，甚至可以直接导出 SVG 和Sketch 文件，我可以理解他支持多件文件格式导出，比如 PPT、word 文档之类的。 然后我还发现一个新的用法，你只要在Midjourny复制sref，就可以生成相同风格的图片了（图4️⃣），比你丢一张参考图出来的效果更加的精准

@sama why gpt 5.5 is not available in my Codex? I'm plus user

feels like a good time to seriously rethink how operating systems and user interfaces are designed (also the internet; there should be a protocol that is equally usable by people and agents)

Honestly thought GPT-5.5 would clearly beat Opus 4.7 on SWE-Bench, but it’s closer than expected. Can't wait to see in practice

@LexnLin What was the prompt to generate thoee imagws

Btw, these were all generated by images 2.0 too :)

@LexnLin sooner than you think

IPhone 20 by GPT images v2

@DavronbekDev Bu darskikmi vibe coding buyicha? Qayerdan topsa buladi

That is why mac is different) Hech qanday kabilarsiz (cabel) ulay olasiz bir biriga

@dilshodhoshimov I feel you😅

pzds

Dude GPT-Image-2... wtf... how.

@bobbyloads 😂😂😂

Visited SuperEgo, hoping for interviews, but left with none. Despite initial hopes, everything appeared normal and under control. No shady business to report. #SuperEgo #Observations

@LexnLin What is deepthink? Give me deepthink link?

DeepThink can oneshot it

Chess and politics always hand in hand

@LexnLin which model is it?

@oncruit DeepThink

left: DeepThink right: GPT5.4 Pro Same prompt btw