pelingpt

AI / LLMs / ML / Vibe Coding

AI Ambassador and Image Gen at @1453ai

[pelingpt @pelingpt]

sürekli dm kutumu aşındıran "kız bu uygulama fikirlerini nereden buluyorsun, ilham perisi mi geliyor" sorularına toplu cevap veriyorum, toplanın sırrımı açıklıyorum 💅🏻

işte benim "vibe coding" sürecimin mutfağı 👇🏻

1. @eachlabs – asıl gizli silahım bu benim için resmen oyun parkı burası. sabah kahvemi alıp direkt buradaki trends ve ai workflows kısmında geziyorum.

olay şu: orada yeni çıkan modelleri (mesela nano banana pro 🍌) veya hazır workflowları inceliyorum. sadece bakmakla kalmıyorsun, test edebiliyorsun. mesela geçen patlayan "kıyafet değiştirme" app fikrini tamamen burada gezerken bulmuştum. "aa bu model çok iyiymiş, bunu hemen app'e çevireyim" diyorum. api işleri falan da bebek işi, entegrasyonu çok kolay. vizyon açmak için birebir.

2. meta & tiktok ads library stalk yeteneğimizi işe çeviriyoruz kızlar. kim reklama deli gibi para harcıyorsa o uygulama tutmuştur.

özellikle tiktok kısmı gen-z trendlerini yakalamak, hangi videonun tuttuğunu görmek için aşırı iyi. rakiplerin copy'sini, görselini analiz edip "ben bunun daha iyisini, daha pembesini yaparım" motivasyonuyla girişiyorum işe.

özetle: ilham gelmesini beklemiyorum, gidip @eachlabs 'ten trendi yakalayıp ads library'de doğruluyorum. bu ikili şaşmaz 🌸

[akira @iso_morph]

I cannot fucking believe that everyone is collectively reinventing all of software engineering from first principles

What the fuck

What happened

[Tibo @thsottiaux]

We have reset usage limits for all Codex and ChatGPT Work users.

Last night around 2am to 4am we suffered an almost global outage. All well and recovered, but you know what comes next.

We learn. We reset. Enjoy.

[Theo - t3.gg @theo]

I got so used to the limits with Fable that Opus feels practically unlimited

[Kenton Varda @KentonVarda]

Building software by prompting agents without ever reading or editing the code directly isn't actually a new thing. It's just called being an engineering *manager*. We have decades of experience about how to be effective as engineering managers. It is a different skill, though.

[Pat Walls @thepatwalls]

Interesting to see @naval getting roasted for showing off his monk life, and @dhh being applauded for showing off his supercars. I don't think that would have been the case 5 years ago.

[each::labs @eachlabs]

🔥One prompt. Four images. Endless possibilities.

From product campaigns and fashion editorials to portraits, menus, interiors, and cinematic scenes.

Here’s what Flint Image can create at 67% lower cost than frontier models 🧵

[Alper Tunga @altudev]

@KorhanErcin I asked Flint to generate quality bar images from Istanbul, and it generated these:

Now I know where you guys trained this model and drank a tasty beer. :)

[Alper Tunga @altudev]

I've been using @eachlabs's model comparison to generate og images for my startups lately, and honestly, it's been great.

Same prompt, 4 different models, 4 different results.

Picked the best one for @formpreneur.

Here's how you can do it too: 👇

[Theo - t3.gg @theo]

There are now 6 labs with a better model than Google

[pelingpt @pelingpt]

@eachlabs eachlabs is cooking smh

[each::labs @eachlabs]

🔥 Something is cooking at each::labs!

One in, four out. Same frontier quality, built for production economics.

Stay tuned 👀

[Peter Yang @petergyang]

Banning Chinese models will be the same self own as banning Chinese EVs.

[Tugser Okur @TugserOkur]

Seedance 2.0

Super casual real smartphone home video footage of a sunny mountain camping day outing with friends. Natural mobile phone camera recording with slight authentic handheld shake, normal frame rate with smooth natural motion, rapid-fire montage with constant quick jump cuts every 1–2 seconds like scrolling through phone memories. Unpolished authentic phone recording of a mixed group laughing, setting up camp, exploring nature, snacking, and taking pictures together. Pure raw home video feel with no cinematic polish or heavy effects.

Use the provided reference photo as the STRICT ONLY visual reference for the main woman. Maintain her exact appearance with zero deviation: [her described features]. Generate a mixed group of friends of all ages around her in a mountain camping environment.

0–2.5s:
Shaky handheld rapid cuts of the main woman laughing with friends near the campsite in the mountains. Light wind gently moves her hair. Quick flashes of hiking boots on dirt and grass, pine trees, and sunlight filtering through the trees.

2.5–5s:
Abrupt jump cuts showing a close-up of her smiling with sunglasses pushed up onto her head, followed by the group setting up a tent, unpacking gear, and taking casual selfies together.

5–7.5s:
Fast handheld footage of her sitting on a camping chair or picnic blanket, chatting animatedly with friends. Snacks, a thermos, and cold drinks are scattered around. Natural sunlight creates subtle lens reflections and realistic exposure changes.

7.5–10s:
Quick cut to a close-up of her giving a big warm smile toward the camera and waving playfully. The footage immediately jumps to the group gathering firewood, arranging camping items, or laughing together near the campfire area.

10–12.5s:
Abrupt edit showing everyone gathered around the campsite, casually passing snacks and drinks around. Friends cheerfully toast with cans or mugs while talking over one another, with mountains and trees visible in the background.

12.5–15s:
Final rapid transition showing the main woman relaxed and sitting among her friends, softly smiling while looking toward the mountain view or forest landscape. The video ends naturally with a gentle phone sway as if someone simply stopped recording.

VISUAL STYLE:

- Natural smartphone video quality.
- Slight realistic handheld shake.
- Smooth normal frame-rate motion.
- Authentic casual interactions and physics.
- Realistic sunlight and exposure adaptation.
- Stable main character consistency.
- Unpolished home phone recording aesthetics.
- No professional stabilization.
- No cinematic color grading.
- No beauty filters.
- No artificial effects.
- No AI artifacts or glitches.

IMPORTANT GENERATION REQUIREMENTS:

- Consistent identity throughout the video.
- Realistic human anatomy and hand interactions.
- Natural walking and body movement.
- Authentic mountain camping ambience and environmental audio.
- Physically correct lighting and shadows.
- Rapid memory-style jump cuts every 1–2 seconds.
- No duplicated people or objects.
- No facial distortions.
- No impossible movements.
- Preserve the casual smartphone home-video feeling from beginning to end.

[Wise @trikcode]

switched my Claude API to Kimi yesterday

it’s faster, 10x cheaper, and writes better code.

i don't care if Beijing knows my db schema.

it's not like my startup has users anyway.

[Anshu @anshuc]

extremely unprofessional. if kimi wants to make it as a frontier lab, they need to act like one: perhaps silently route people to worse models, and maybe write a blog post about the collapse of humanity

[David Sacks @DavidSacks]

Kimi K3 just fixed 15 critical security bugs that Codex and Fable refused because of “cyber guardrails.” There’s no reason to limit American models on tasks that Chinese models handle without issue. We’re only making ourselves less competitive.

[each::labs @eachlabs]

🌍 Want to create a 15-second country vlog with AI?

In 3 simple steps:

1- Upload your photo
2- Choose any country
3- Generate your vlog with Seedance 2.0

Your travel vlog is ready. ✨

Workflow below 👇

[Alper Tunga @altudev]

Most US LLC services assume you have a team, a lawyer, and money to burn.

Solopreneurs don't.

That's why I built Formpreneur for non-US solo founders.

It's a founder-to-founder app with clear pricing from $329.

It's all included. State filing + registered agent. EIN + company docs. 1-on-1 Banking and Stripe setup help.

No upsells, no add-on maze.

Built this because I needed it back then. Maybe you do too.

https://t.co/tHzuCNMdnL

[Qwen @Alibaba_Qwen]

Qwen3.8 is launching and going open-weight soon!🌐

With a massive 2.4T parameters, this model is continuously evolving. We believe it’s one of the most powerful model available today, compatible to leading frontier AI models , second only to Fable 5.

You don't have to wait to test it. Just now, the Qwen3.8-Max-Preview made its debut on Alibaba’s Token Plan, Qoder, and QoderWork. Be among the very first to try it out.

Can't wait to hear what you build. Stay tuned! 🚀 

Token Plan
international：https://t.co/YRvcGdB9Bv
China：https://t.co/PKMUNwUuRp

[Tugser Okur @TugserOkur]

🚨 Google sessizce çok güçlü bir yapay zekayı serbest bıraktı.

Adı: TimesFM

Satış tahmini, hisse senedi fiyatı, web trafiği, enerji tüketimi, kripto oynaklığı…

Tüm bunları sıfır eğitimle, direkt tahmin edebiliyor.

Özellikleri:

→ 100 milyar gerçek veriyle eğitildi
→ Zero-shot çalışıyor (hiç fine-tune etmeden)
→ Tamamen yerel cihazında çalışıyor
→ %100 ücretsiz ve açık kaynak

Bu model, tahmin işini yeniden tanımlıyor.

Link aşağıda 📷