@ilqaur

someone

Joined October 2013
dude retweeted
Learn to fine tune a model this year Learn to fine tune a model this year Learn to fine tune a model this year Learn to fine tune a model this year Learn to fine tune a model this year Start small Start slow But make sure you start
Fine-tuning is about to become one of the most valuable AI engineering skills. Not because everyone needs a custom model. But because the people who understand how models learn from data will build things others can’t. Full guide:
58
461
17
6,129
819,223
been thinking about systems a lot lately starting to turn that into something very early, very wip
11
17
1
677
21,337
اگر میخواید دیزاین فیگما بدید به Agentتون کدش رو بزنه، اسکرین شات گرفتن خیلی کار رو دقیق در نمیاره و MCP خود فیگما هم کلا ۶ تا ریکوئست در ماه(!)‌ میذاره بزنید و بعد میگه پول بده. با این mcp میتونید دورش بزنید و دیگه به API Limit فیگما نمیخورید: github.com/gethopp/figma-mcp…
2
7
1
162
6,028
یکی از کتاب‌هایی که سال‌ها پیش خوندم — و دیدم رو نسبت به عدد و رقم و شهودی که روزانه ازش استفاده می‌کنیم — عوض کرد The power of mathematical thinking بود. تو خیلی بحث‌ها مثال‌های اشتباه‌های شایع کتاب رو می‌بینم مخصوصا وقتی بحث عدد و رقم و اکثریت و اقلیته. خلاصه کتاب خیلی خوبیه و خودم هم دوباره باید بخونمش
1
2
43
2,363
dude retweeted
you have no idea how far this goes. This is just the tip of the iceberg
Mentally imagining muscle contractions increased strength without exercise.
166
2,346
45
34,661
1,421,418
Update from CENTCOM Commander:
1,272
3,119
339
18,658
2,218,240
If you were wondering why I vanished between 2022-25, it was because I was getting absolutely wrecked by this project (◡‿◡✿) still worth it notion.com/blog/how-notion-h…
3
14
2
243
8,372
به به عجب پاییزی بشه 🤌🤌 AI for Games
1
21
228
This is genius. > HarnessRouter. A router not for models, but for harnesses.
14
13
5
172
14,231
Harvard's "Dynamical Systems" Lecture Notes: abel.math.harvard.edu/~knill…
74
475
26,284
My model + harness stack these days: 1. Kimi K3 is the best OSS coding model. GLM-5.3 is next. 2. DeepSeek V4.1 Flash is my pick for easy/moderate tasks: insanely fast and cheap. 3. I reach for Astra for complex tasks. 4. Codex works better with OSS models. Claude Code can make the same model cost 2x more.
108
57
10
1,329
75,919
new post: the senior engineer death spiral sunilpai.dev/posts/the-senio… a friend just started a big job and asked for some advice. so I braindumped a monologue about a super common failure mode I see with engineers and posted it here, hope it helps whoever it can.
145
236
95
3,369
460,765
A French military source told Le Monde that France is already carrying out undisclosed conventional military responses to Russian hybrid attacks. Asked whether Paris could retaliate militarily without making it public, the source replied: “That is already happening.”
80
1,036
168
8,524
758,394
Jev by @typesafeai is free on Vercel AI Gateway until Sept 25. Build with the fastest adopted model on the Gateway at no cost. vercel.com/ai-gateway/models… nitter.cf/vercel/status/21010773…
Jev was adopted faster than any other model in AI Gateway history. In the first day, @typesafeai reached ~13% of teams, 2x the GPT-5.6 family and 6x Fable 5.1.
73
231
133
2,326
689,162
Jev has completely replaced ChatGPT, Codex, Cursor, Claude Code, iMessage, Spotify, Ford, CVS, and McDonalds for me. It really is the everything model.
Theo's opinion on Jev (Type Safe AI)
157
52
25
2,153
335,578
WalShadow is probably one of the most ambitious projects we’ve taken on. When we decided to build this six months ago, we genuinely weren’t sure how it would pan out. We were going directly after Postgres’s physical write-ahead log, an area backed by decades of deep engineering expertise. Now, seeing the early results with customers, especially Postgres → ClickHouse replication latency in the hundreds of milliseconds, it feels totally worth it. It opened up a new world tousers, that wasn’t possible before. This wouldn’t have been possible without the team’s deep expertise in Postgres replication, the operational lessons we’ve learned onboarding thousands of customers to Postgres CDC, and our experience deep in the Postgres replication and backups, creating and contributing to WAL-G and WAL-RUS I’m really excited for more customers to try it out. And perhaps the most exciting part: this brings ClickHouse much closer to becoming a physical analytics standby for Postgres. 🚀 github.com/ClickHouse/walsha… @ClickHouseDB
2
15
155
7,765
dude retweeted
Notion runs one of the largest CRDT deployments in production, processing millions of operations every minute! We wrote about the data structures we adopted from CRDT research to make our text editor more collaborative, and how we built on those concepts to fit our block-based data model. notion.com/blog/how-notion-h…
31
59
8
490
34,158
روزی ۸-۹ ساعت رو اپ دارم وقت میزارم. و کم کم داره تبدیل میشه به چیزی که میخوام. حیف ویندوزی ها هنوز امکان استفاده ازش رو ندارن (فعلا رو لینوکسه)
48
9
1
375
15,380
This is why I am on Twitter, to see posts like this
Solving a Rubik's Cube with graph theory.
156
15,391
86
195,324
5,382,938