constantin (@luckenco) header
constantin (@luckenco) profile picture

Head of Prosperity ⌯ Herding silicon intelligence Implementing AI for DACH companies @ https://teapawt.org/t.co/fjuW8xPUY8

Tübingen, Deutschland
Joined January 2025
If you are into short and intense workouts, you might like repsquid. App Store link is in the first comment below.
10
7
6
55
11,151
new perk unlocked: “back massage”
8
277
training hard for honorary unc status
2
23
560
constantin retweeted
yapatron in hong kong
3
1
15
470
BREAKING: Famous Austrian restaurant critic @parzerp was recently spotted in Hong Kong 🚨
1
8
295
> you’re quite tall and not fat - thats quite good > hmm hard to find pants tight at the waist because uhh model figure asian store lady knows how to sauce a white boy up (I bought the pants)
1
11
366
I got blessed with a Chinese unc sleeping the whole long-haul flight.
2
8
214
completely unrelated to recent events
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. The world needs both frontier closed models and frontier open models. images.nvidia.com/pdf/Open-W…
1
6
159
constantin retweeted
UX latency is the most frustrating (lack of) advancement in computing over the last ~50 years. IBM, way back in the 80s, found that user productivity shot up *exponentially* as system response time dropped. The "Doherty Threshold" was defined as ~400 milliseconds; the maximum amount of tolerable UX latency before interrupting user flow state. It's unbelievable how many modern operating systems (*cough* win11) violate this threshold constantly. Literally all you need to do as a Dev is pay attention. Add a "Doherty Threshold" to your test suite. There's so many different profiling tools, and especially with how good models are, really no excuse. Sure, some stalls are unavoidable, but that's when you pull out other UX tricks to at least give the illusion of responsiveness. Crazy how much vintage computing hardware feels immeasurably more responsive despite having thousands of times less overall performance.
198
567
59
6,452
151,199
constantin retweeted
Qwen4 is coming. Deeper knowledge. Richer world understanding. Stronger coding and agentic capabilities. Built for real-world tasks — from complex coding and knowledge work to front-end builds and professional software workflows. Qwen4-Max is now in preview on Alibaba’s Token Plan. You can also get early access to the new model on QwenWork, Qoder, Qwen APP, and QwenStudio. Put it to the test on a real problem. Show us what you build. 👀 Token Plan: qwencloud.com/pricing/token-…
986
1,540
1,073
12,734
1,402,621
claude != ai
5
66
tired: slide to merge wired: hold to debrief
“LGTM,” he whispered from the toilet as the merge slider glided beneath his thumb so smoothly that it briefly suspended the laws of friction.
1
1
4
450
third tester of [redacted] is being onboarded as we speak. Billions must kode
3
204
Increasing model capabilities decrease the need to be at your desk. I respect and agree with the direction, but the execution feels off (not linear but most of the phone apps) Mobile only is valid if you believe that model improvements are going to continue. But then there is no need for diffs to be a first-class citizen. No need for most of the patterns and UI we currently use. We are switching paradigms but are taking all the primitives of the past with us? Let’s be a bit more creative.
Introducing Mobile Diffs. • Understand changes with Guided Reviews • Start on mobile, continue on desktop • See what needs review in your inbox • View preview links • Reply to diff discussions • Iterate with coding agents • Merge approved changes All from your phone.
2
1
4
306
constantin retweeted
*sshole is out if you run sshd on alpine, openwrt or any musl box with password auth on: update now official writeup: sshhole.com
86552c7e1e5d46882844cffabb44d454129ff22927f13797b151c9db4519d47d
188
890
446
5,449
1,572,925
broken Chinese is coming from the living room.. the mrs is currently exercising her bargaining skills in mandarin. the adversary is grok, the battlefield seems to be a the price of a blouse. we are going to scoop up some banger deals on the China trip
18
361
that’s not a joke btw “FDEs” (everyone) should see more, do more and think more. there is 0 reason for you to be glued to a screen inside all day
if your coding environment isn’t optimized for walking around all day you’re missing out
1
19
628
constantin retweeted
So much value in the skill of: - Grab a book - Turn off distractions - Focus, read, and learn
33
68
9
1,303
45,137
agents are super horny to turn the light green. they want to pass the test. they also have a complete disregard for non-functional properties.
3
19
724
really interested how that turns out. it's basically this but newer teapawt.org/luckenco/status/199267…
the team is experimenting this week at switching all our subagents to deepseek flash and using opus 5.5 only on the primary subagent
4
360
post-ai developers (myself included) forget that code reviews were also in place to spread knowledge and diffuse understanding of the problems the software is trying to solve. it's not just about catching bugs, models are quite good at that. adversarial review's effectiveness is more and more questionable. with smaller and smaller teams (due to ai?) this might be less of a problem, but so far I have no adequate solution for this.
3
13
332