fafa.👩🏻💻 retweeted
The Rise Of AI Coworkers
The bottleneck has moved from writing code to setting goals, so the winning setup lets agents check their own work while humans own taste
youtube.com/watch?v=y1-9yeYL…
fafa.👩🏻💻 retweeted
The Future of Data Science and Analytics Engineering at Klaviyo
When agents write the code, the scarce resource becomes human review, so data science's job turns into setting constraints and proving the work held
klaviyo.tech/the-future-of-d…
fafa.👩🏻💻 retweeted
Will AI-native services earn software margins, or stay staffing firms with better software?
fafa.👩🏻💻 retweeted
Like a leadership team with unlimited consultants, where the edge is knowing your own operation well enough to ask the right question.
fafa.👩🏻💻 retweeted
Claude Dashboards and the shift from chart builder to chart specifier
thedatavist.substack.com/p/c…
The scraper's licence is cheap. Its maintainer is not.
⏱️ Price the engineer-hours Firecrawl gives back
📊 Fewer broken jobs, fewer 2am fixes
🧠 Spend the hours on the product, not the fetcher
firecrawl.link/s1IFcUo
24
Tencent Launches Hy4 flash to Power High-Speed, Cost-Efficient AI Deployment for Everyday Workflows
tencent.com/tencent-launches…
53
Claude was trained on protein sequences at scale xinmingtu.cn/blog/2026/claud…
52
fafa.👩🏻💻 retweeted
Engineers naturally want to encode their own clever ideas and domain insights into systems design. That's provides personal satisfaction and better outcomes today, but will soon be inferior to letting agents discover the right abstractions themselves
Research note: Measuring alignment eval tells metr.org/notes/2026-11-16-me…
70
19k enrolled in GenAI for devs. Go past autocomplete at work.
⭐ 4.7 from 11.8K+ reviews
🛠️ Code review, debugging, agentic workflows
💻 3 courses covering DevSecOps, CI/CD
imp.i384100.net/jroKRn
110
fafa.👩🏻💻 retweeted
I integrated Jev model by @typesafeai as autonomous tram driver.
Not real time video, it processes existing tram drive route and suggests actions.
Very cool.
fafa.👩🏻💻 retweeted
This is a good read from Anthropic on why you don't really need subagents anymore, and most workflows should just be one agent. Bitter lesson adjacent. Lots of good bits in here, both why this is true, and how you can convert back to a single agent architecture.
Some quick notes on what I found interesting:
- Modeling your agents around your org chart is *not* what you should do
- Plan mode is dead. Better to have the plan be in the main agent's context
- MCP tool design is critical. Take a jobs to be done approach versus mapping your API to tools directly.
- Build evaluations on the full task completion / final output (or on the skills / tool calls if you need to). Whatever the case, make sure evals are coherent and non-conflicting
- If you do keep subagents around, call it as a tool so that the main agent controls the context
Also thematic in this post: Many attempts at AI adoption layer it on the existing architecture (both technical and organizational), when the most appropriate way to adopt it is to re-think the architecture completely. This is most difficult on the people side (e.g org design, what teams should do, what teams should even exist, etc.)
Def. Worth reading the entire thing if you are doing applied AI stuff.
claude.com/resources/article…
دیگه برای ساخت مدلهای 3D لازم نیست Blender بلد باشی! 😍
تیم رسمی Blender، فرآیند مدلسازی، ساخت متریال، نورپردازی و رندر رو به ۲۸ تا Skill تبدیل کرده.
کافیه این مجموعه رو نصب کنی و در Claude Code یا Codex با زبان طبیعی توضیح بدی چه صحنه یا مدلی میخوای.
هوش مصنوعی خودش با استفاده از اسکریپتهای Blender:
- مدل سهبعدی رو میسازه.
- متریالها رو تنظیم میکنه.
- نورپردازی رو انجام میده.
- در نهایت از صحنه خروجی رندر میگیره.
نکته جالب اینه که خروجی فقط یک تصویر نیست!
مدلهای ساختهشده همچنان قابل ویرایش و بررسی هستن، میتونی ایرادهاشون رو برطرف کنی و حتی بهعنوان Game Asset خروجی بگیری.
این پروژه متنباز و رایگانه.
فقط به Blender 5.2 یا بالاتر نیاز داری؛ حتی میتونی نصب Blender رو هم به Codex بسپری!
🔗 گیتهاب پروژه:
github.com/EL4CTEO/blender-s…
کانال تلگرام : t.me/yarad_ai
> be poteto
> real name: Lauren Tan
> discover design at thirteen
> get paid to redesign your school website
> study finance, fall in love with coding
> start a business, run out of money
> one conference talk opens the DockYard door
> move from Australia to America
> nearly four years at Netflix
> six years at Meta, then join Cursor
> Cursor gets bought by Elon Musk's SpaceX
> now help build Grok Bot
she spent years at Netflix and Meta, now give her one hour to change how you work with AI ↓
Your coding agent's bill should not outgrow the code it ships.
💰 GLM-5.3 covers the daily agent loop
🧠 Reasoning, coding and agents in one sub
🛠️ Prompt caching cuts the cost of repeats
go.sjv.io/k4oxzM
fafa.👩🏻💻 retweeted
Understanding system design helps you build better apps, even when vibe coding.
Better architecture decisions. Better AI prompts. Fewer surprises when your app grows.
So I built an interactive app to learn system design visually.
designed-to-scale.vercel.app
fafa.👩🏻💻 retweeted
The CEO’s singular impact on the success or failure of AI in organisations
mckinsey.com/capabilities/st…
fafa.👩🏻💻 retweeted
Scaling AI-Native Analytics Patterns at Meta
Standardise your measurement patterns and document them well, because agents can only be reliable on top of a governed foundation
medium.com/@AnalyticsAtMeta/…














