Sabitlenmiş Tweet
Jukka
1.6K posts
Jukka
@jugibuilds
Cooking an agent orchestration platform • I use Claude Code 12 hours a day
Katılım Ocak 2012
1.2K Takip Edilen826 Takipçiler
I don’t read AI generated code.
I don’t read AI generated tests.
If AI says everything is good, I ship to production.
This is a flawless strategy.
Moses Xu@mosesxu
@svpino i don't read every line CC writes for me. i define what needs to be tested and CC writes the tests too. if they pass it ships. "i understand every line in my codebase" was already a lie before AI showed up, now we just can't pretend anymore
English
Applies to new features too not just first launches.
Better to get users to use the stuff and spot obvious bugs.
When they report something it tells you they care and they too want to get it working.
Also a signal to build more.
And a great opportunity to get to know the users.
English
I failed with StorySDK
Spent a year building features. Waiting for the “perfect” launch. Then I finally launched… And no one used those features. Not even once.
But users started messaging me: asking for things I never even considered. They had problems to solve. I was thinking in features.
Lesson: Start with the problem. Features can come later.
Don’t delay the launch.
Viktor Seraleev@seraleev
A lot of developers obsess about features Users rarely care. They care about solving one small problem quickly
English
Took me roughly 6 months of intensive daily work with AI coding agents to really get the imagination moving.
Not sure if it's possible to alter thinking in the required way with just low-intensity AI exposure.
If you're already behind, get your shit together now.
Fuli Luo@_LuoFuli
Imagination is the ceiling of productivity in the new era. Inspiring imagination is the core of management in the age of Claw.
English
@zach_yadegari how do you estimate/calculate LTV for new businesses? pretty easy to just believe that yes our users stay 3 years
English
It's really interesting how so few people truly understand how distribution works. There is no magic sauce. If it's a sole-user experience, you just need to learn how CAC : LTV works and then how to operate social media platforms on both the organic and paid side. If it's a social experience, you need to learn how k-factor works (much harder). Stop doing stupid shit like throwing flyers out of airplanes.
English
was messing with the OpenAI base URL in Cursor and caught this
accounts/anysphere/models/kimi-k2p5-rl-0317-s515-fast
so composer 2 is just Kimi K2.5 with RL
at least rename the model ID
Cursor@cursor_ai
Composer 2 is now available in Cursor.
English
the biggest AI opportunity right now isn't building agents
it's understanding businesses well enough to build the RIGHT agent
marketing agents specifically
every business needs one... almost nobody is building them properly
what i see instead:
> generic "content teams" that pump out forgettable posts
> bullshit automations that feel like spam
> systems built around AI capabilities instead of business needs
flip that
start with the business problem, map the funnel, understand the audience... then build a system that fits like a glove
the AI part is honestly the easy part, understanding businesses is harder
English
Fluar just got acquired in an all cash, 6-figure deal
I built it solo for 14 months, now it’s going to an industry leading team that can really scale it
this is the second startup I’ve sold since I learned to code 3 years ago
life is incredible, what a time to be alive 🫡 LFGG
English
@tminusdev @fynnso Yeah they add some magic on top of models.
It’s actually a great way to monetize deep knowledge that others don’t (yet) have.
Modern consulting.
English
@jugibuilds @fynnso Probably the system prompt, which is arguably better than GitHub copilot
English
Jukka retweetledi
@jugibuilds @fynnso If there is no moat why don't you walk across?
English
Measuring token burn is the same as measuring lines of code (more = better).
Zero is bad but an insanely large number is also bad.
sunny madra@sundeep
“If your $500K engineer isn’t burning at least $250K in tokens, something is wrong.”
English
Let me tell you a story.
When I launched my first app, it started growing purely on ASO. No ads. Just organic.
Then something happened.
A competitor showed up. First: they used incentivized traffic to climb to the top. Then: they turned on ads.
And just like that… I lost my traffic.
That was the moment I learned: ASO alone is not a strategy. It’s a vulnerability.
If you rely only on organic, sooner or later someone will come in, spend money, and take your position.
Since then, I always combine ASO + paid acquisition.
Because ASO has another problem – it’s capped. There’s only so much traffic you can get.
If you want to scale, you need ads. You need social. You need distribution outside the store.
Otherwise you’re just waiting for someone to outbid you.
Thomas Burkhart 💙@ThomasBurkhartB
@seraleev Sorry that's wrong. I saw clear growth by better ASO
English
I don’t know what you call them, but these little side tabs are like the emdash of vibe coded UIs
Juri Strumpflohner@juristr
What's your AI adoption level? (according to Steve Yegge)
English
One underrated skill for mobile developers:
Understanding user psychology.
Why people install.
Why they subscribe.
Why they churn.
English
Custom code factories is the way to go. QA is becoming the bottleneck though.
The mega productivity unlock is to build an autonomous custom pipeline that wraps a coding agent to do things the way your company wants.
I just last week tripled my team's output by developing a custom Docker based agent.
The container runs Claude Code with the exact stack we use and completes long running tasks with zero babysitting from start to finish.
Exciting times.
English
Agentic software engineering adoption is on fire at @Uber. 1,800 code changes per week are now written entirely by Uber's internal background coding agent, and 95% of our engineers now use AI every month across all the tools we track.
This is a real reset moment for engineering; it's one of the most exciting times to lead. This shift requires builders to be curious and hands-on. I’m incredibly lucky to be surrounded by a team that’s doing exactly that.
The best part is that the strongest adoption isn’t being pushed top down from leadership announcements; it’s coming from engineers who are quietly experimenting, quietly shipping, and quietly pushing things forward.
I love spending time with those engineers because there’s no substitute for being close to the work.
Over the last few months, we leaned in hard, and the results have been phenomenal.
The bigger shift: going agentic.
84% of AI users are now working with agent-style workflows, not just tab completion. Claude Code usage nearly doubled in 2 months (32% → 63%), while IDE-based tools have largely plateaued.
Engineers are moving from accepting suggestions to delegating tasks. Even within traditional IDEs, ~70% of committed code is now AI-generated.
Background agents are writing code autonomously.
Our internal background coding agent went from <1% of all code changes to 8% in just a few months. There is zero human authoring. Engineers review and approve, but the code is written entirely by AI agents.
The role of the engineer is shifting - from writing every line to architecting systems and reviewing AI-generated code.
More to come from the @UberEng team in the coming days.
English
who gives a fuck about code formatting in 2026 lol
daily.dev@dailydotdev
oxfmt is 30x faster than prettier. most devs will stick with prettier because we're tired of switching tools every 3 months for nothing
English