λL-D1 | AI for Buzzer 🍉 (@F2aldi) / X

λL-D1 | AI for Buzzer 🍉

88.7K posts

λL-D1 | AI for Buzzer 🍉

@F2aldi

Share Tips AI, SWE and Work Culture | Business Inq on [email protected] | @cursor_ai ambassador | @open_ai codex ambassador | @JatevoId Contributor

Jakarta Selatan, DKI Jakarta

Joined December 2011

Replies Articles Media

Pinned
λL-D1 | AI for Buzzer 🍉
@F2aldi
Jul 28
Kalau datang ke AWS Summit Jakarta bareng tim, usahakan ikut sesi yang berbeda-beda. Soalnya, implementasi AI perlu kerja sama semua tim, bukan engineer saja. Developer kudu ngerti masalah teknis. Tim bisnis dan leader perlu memahami impact, governance, dan cost.
00:00
Paid partnership
λL-D1 | AI for Buzzer 🍉
@F2aldi
9m
Ini hot take-ku ya, jasa review CV itu bener2 abu-abu. Kenapa? - HR itu manusia, pasti ada bias. Manusia A bisa anggap CV B bagus, tapi manusia C bisa anggap CV B jelek :) - Keterima kerja itu enggak hanya perkara “cara menulis CV” aja. Ada “isi CV”-nya, relevansi pengalaman
WORK
@worksfess
3h
Automated by @magnitvde
Ada yg pernah pake jasa review cv? Mending pake jasa itu atau pake AI guys? Work!
λL-D1 | AI for Buzzer 🍉
@F2aldi
17h
Bisa, khususnya kalau pakai Codex. Untuk menyamakan workflow satu tim, taruh konfigurasi langsung di repository: AGENTSmd untuk project instructions dan aturan kerja .agents/skills/ untuk reusable workflow, references, templates, dan scripts .codex/config.toml atau
AWP 🇵🇸
@adith_wp
19h
Buat yang make Claude Code or Codex dll tapi as a team (di company). Gimana cara kalian nge-enforce workflow? Kan susah tuh mastiin semua punya workflow yang sama even di satu repo. Belum lagi kalo reponya banyak (misal 50 lebih).
λL-D1 | AI for Buzzer 🍉
@F2aldi
21h
Speechless...
λL-D1 | AI for Buzzer 🍉
@F2aldi
Jul 30
TL;DR: - GPT-5.6 Sol awalnya score 13,3% pakai official harness. - OpenAI mengubah 2 settings: reasoning tetap disimpan dan context lama di-compact, bukan dihapus. - Modelnya sama, cuman sistem memory dan harness-nya yang beda. - Score-nya naik jadi 38,3% dengan output token 6×
Tibo
@thsottiaux
Jul 29
Turns out GPT-5.6 Sol is actually SoTA on ARC-AGI-3. Just took two setting changes. You just have to allow it to reason and work over multiple context windows with the help of our canonical compaction implementation. openai.com/index/how-two-…
See @F2aldi's full profile
Sign up
Log in

See @F2aldi's full profile

See @F2aldi's full profile