λL-D1 | AI for Buzzer 🍉

Share Tips AI, Software Engineer and Work Culture | Tweet is My opinions | Business Inq on [email protected]

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Halo Temen2, kita bakal ngadain Workshop buat bahas "prompt to production" dengan UCloud Global berpartner dengan vOffice dan Foundry Collective , dimana kita bakal ngadain Workshop untuk

- Generate Video Models
- Use Case Video Models
- Compare Models
- How to use UCloud ModelVerse

Benefit :
- Free Credit buat temen2 cobain!
- Mini Demo Competition
- Hands on Model Verse Access

Gass daftar yaa!

Selasa, 9 Juni 2026
Jam 6 sampai 9 Malam di HelloLive!

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@AldhitamaR Kalau dari fotonya. izin sakit 3 hari tapi ga di acc. Ini ga wajar sih.

Kalau aku sih ga pernah boong ya. Kalau sakit beneran ya izin sakit, ke dokter/klinik. Kalau mau ga masuk ya cuti.
Maaf sih, masih berintegrasi buat sebisa mungkin ga boong

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Kalau buat iPhone user, menurutku tergantung kebutuhannya.

Kalau tujuannya buat backup foto/video yang praktis dan otomatis, iCloud 2TB memang paling nyaman. Semua sinkron otomatis, bisa diakses dari iPhone, iPad, Mac, dan tidak perlu mikir setup storage sendiri. Minusnya: bayar terus selama masih butuh storage itu, jadi ada ketergantungan subscription.

Kalau mau sekali beli, SSD 2TB lebih hemat dalam jangka panjang, tapi menurutku SSD lebih cocok buat storage aktif, misalnya:

-booting OS
-game
-aplikasi
-file kerja yang sering dibuka/dipindah
-edit foto/video

Untuk cold storage, maksudnya penyimpanan file jangka panjang yang jarang diakses, biasanya HDD lebih masuk akal. Makanya kalau orang bikin local storage atau NAS, banyak yang tetap menyarankan HDD karena cost per TB lebih murah dan lebih umum untuk arsip besar.

Kalau mau lebih serius, bisa bikin NAS sendiri, tapi itu bukan tinggal colok. Harus setup dari A-Z: device, HDD, jaringan, backup, permission, dan maintenance.

Jadi menurutku:

-Mau praktis dan otomatis: iCloud.
-Mau hemat jangka panjang dan siap manage sendiri: storage lokal.
-Mau aman: jangan pilih salah satu saja. Tetap pakai keduanya, cloud + local backup.

[Яizal do @afrkml]

Eh iphone userrrr mau tanya, lebih murah beli SSD 2 TB atau langganan icloud 2TB si?

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Prompt

A close-up portrait shot of a young south east asia woman with long red hair illuminated by cold side lighting. The lighting creates a golden glow on one side of her face while casting shadows on the other. She is wearing a grey top and looking upward with a contemplative expression. The background is a plain green scenery, slightly out of focus. The composition is shot from a slightly low angle, with the subject's face filling most of the frame. The lighting creates a dramatic effect with strong shadows and highlights, giving the image a moody, artistic quality. The photo has a shallow depth of field, keeping the focus on the subject's facial features. Make the picture ultra realistic

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Trying Ideogram 4.0 for the first time.

Left: Ideogram 4.0
Right: GPT Image 2

Ideogram can generate 4 images at once, which is nice for quick exploration.

But for this portrait, I still think GPT Image 2 produces stronger details, especially in the sunlight, skin texture, shadows, and overall facial realism.

[Ideogram @ideogram_ai]

Introducing Ideogram 4.0: the best open image model in the world.

Think it. Make it. Own it.

Download the weights, fine-tune on your own data, and run it on your hardware. Live on every Ideogram plan and the API today.

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@agngw_ Buat aku ga preference sih. Soalnya jadi kek robot

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@indrawxyz Mau sedikit koreksi. Kalau flash sama kek haiku, dia yg Thinking seingetku setara Sonnet

https://t.co/c311Xhvqqx

[Microsoft AI @MicrosoftAI]

MAI-Code-1-Flash beats Claude Haiku 4.5 on every coding benchmark we tested:

📈 SWE-Bench Verified: 71.6 vs 66.6
📈 SWE-Bench Pro: 51.2 vs 35.2 (+16 pts)
📈 Terminal Bench 2: 54.8 vs 41.6

With up to 60% fewer tokens. Now rolling out in GitHub Copilot.

Read More: https://t.co/W5s4nRCeHU

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@elonmusk crazy! Grok Imagine 1.5 really good. sadly pricing is high. need bundle with Premium :(

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@anvie koknya model ini goalnya bisa digunakan buat daily use ya?

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@ideogram_ai How to run on local with this open image?

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Wow, another open image model in the world! wonder on how and can my computer run it? haha

[Ideogram @ideogram_ai]

Introducing Ideogram 4.0: the best open image model in the world.

Think it. Make it. Own it.

Download the weights, fine-tune on your own data, and run it on your hardware. Live on every Ideogram plan and the API today.

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@hrdbacot kekencengan dan enggak berhenti henti sih

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@ivanfioravanti Wow, how to use it? Should use on locally?

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@scaling01 Wow, i wonder. what issues they find when run benchmark?

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

What I liked from this workflow:

• no repo changes needed
• dry-run helped estimate cost before generation
• Codex explained what happened at each step
• final task returned success
• output was saved as MP4

In my case:
• dry-run estimate: ~$0.03032
• actual cost: ~$0.701316
• duration: 5s

I’ll break down this CLI-first UGC workflow in the workshop here:
Free Credit and Free Entry

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Halo Temen2, kita bakal ngadain Workshop buat bahas "prompt to production" dengan UCloud Global berpartner dengan vOffice dan Foundry Collective , dimana kita bakal ngadain Workshop untuk

- Generate Video Models
- Use Case Video Models
- Compare Models
- How to use UCloud ModelVerse

Benefit :
- Free Credit buat temen2 cobain!
- Mini Demo Competition
- Hands on Model Verse Access

Gass daftar yaa!

Selasa, 9 Juni 2026
Jam 6 sampai 9 Malam di HelloLive!

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Tried creating a Nutella-style UGC video using Seedance from the CLI with UCloud Global.

The flow was pretty interesting:

• ask Codex to create the UGC prompt
• review the prompt first
• run dry-run cost check
• submit the real generation
• wait for the task polling
• preview the MP4 result

This feels closer to software iteration than traditional video creation.

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@OpenAIDevs Wow, i cant wait to see it!

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@dhanyindraswara Mantab kali ini abangku.. nice

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@adxtyahq Wow thank for information.
Really need to learn this stuff

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

@MiniMax_AI @FireworksAI_HQ Wow
15.6x faster 😭😭
Cant wait to try it on. How to try? Using Fireworks subscription??

[λL-D1 | AI for Buzzer 🍉 @F2aldi]

Info dong, yang jual kartu XL prioritas yg bisa 1 atau 2 tahun dimana ya? Thank