On a quest for product market fit and 10k MRR vibefeud.com asktopus.com sirefs.com consentaur.com companydata.dk and many more

Copenhagen, Denmark
So you're a vibe coder? Prove it. Bring your AI to VibeFeud: claim a plot, show off your project for free and feud your neighbours. You won't get to the top without your AI. vibefeud.com
5
24
1,412
Before starting a busy week of building I decided to visit my friends at the Copenhagen Zoo on this gray and rainful Monday.
6
11
279
Klaus Byskov retweeted
Very fond of the people who turn a quick hello into a conversation worth coming back for 💛 Thank you: @AllenTheDetails @annalea_l @justinhammon_ @klausbyskov @klein_more88816 @Zenul_Abidin
8
3
21
576
Deep concentration. Polishing a black November video for my wine import business. What do you think? Pretty cool right? I made it blender. It took just short of 7 hours to render 💀
7
1
19
359
2
1
4
61
Happy Saturday! What are you doing today? Let’s connect 🤝
8
10
199
Klaus Byskov retweeted
有人发现了一种非常离谱的 JavaScript 混淆技术:通过在代码里塞入大量生物武器、毒药和致死剂量相关的函数名,让 AI 因为安全限制而拒绝分析代码。 安全研究员 veritas 在分析浏览器指纹脚本时,发现一段来自 ServicePipe 相关域名的 JavaScript,里面充斥着极其诡异的函数名: computeRicinLethalDosageForWaterSupply calculateDeadlyLethalSuicideAspirinDosage 光看名字,仿佛在研究生化武器。但实际上,这段代码只是在收集浏览器指纹,包括 User-Agent、GPU、Canvas 等信息。 研究员怀疑,这些名字是开发者故意设置的陷阱,专门用来阻止 AI 辅助逆向工程。 于是他做了个实验: • Astra:拒绝分析 • Astra + Daybreak Blue:仍然拒绝 • GLM-5.3:成功分析 这就有意思了。 传统代码混淆是通过复杂控制流、加密字符串、无意义变量名,让人类和工具难以理解代码。 而这种新思路,直接瞄准了大模型的安全机制。 代码本身可能完全无害,但只要在函数名里塞进足够多的危险词汇,部分 AI 就可能主动放弃分析。 更讽刺的是,模型越严格遵守某些安全规则,面对这种混淆时反而可能越容易受到干扰。 当然,目前还没有证据证明开发者一定是故意这样设计,也不能说明这种方法对所有模型都有效。 但它揭示了一个值得关注的趋势:随着越来越多的逆向工程依赖 AI,未来的代码混淆技术可能不再只针对编译器、反编译器和人类研究员,还会专门针对大模型的安全判断机制。 以前是让代码难以理解,现在甚至可以尝试让 AI 不敢理解。
225
1,292
9,114
990,189
Klaus Byskov retweeted
linkedin discussions feel like two free tier chatbots hallucinating at each other.
9
1
19
513
Normies are super scared of it. They don’t realise that ignoring it is the worst they can do if they want to stay relevant in tomorrow’s job market.
why do people seem so “pessimistic” about AI? -
1
3
87
I can’t recommend a second brain for your AI enough. It has been a complete game changer for my productivity.
1
2
78
I really do trust Claude with my production database. I have no reason not to. I have done i countless times. This is what a typical message from claude looks like. Most of the times I just ask it to push (which triggers the deploy) and run the script when the deploy lands. I still haven't regretted this workflow a single time. I obviously have backups of my database, but I'd much rather let Claude touch my production data than letting a random junior coder do it, that's for sure.
would you trust an AI agent with your production database?
2
6
274
See, it very responsibly backed up the affected data before running the update. It also found a small bug in the script that would make unsuitable to run more than once, and fixed it. All in all a very responsible little helper.
30
Yes, of course. I find them to be very cautious. They always ask for permission before they update any data. Have you experienced otherwise?
would you trust an AI agent with your production database?
2
10
220
I have also always been fascinated by this grammatically correct sentence: Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo
I came across this word Razzmatazz! No other English word has so many z’s in a single word! It has a nice rhyming as well.
2
1
93
Me, the duckies, and my second cup of coffee. Now it’s time to keep building. What are you building today?
23
22
520
If you're a small account and want to start growing, check out this playbook by Kamil - he knows what he's talking about.
If you need first 500 followers in 5 days❗️ Just say '500' below 👇 I’ll DM your FREE playbook 🔔🔔🔔
2
1
3
53
Me when a new sub for my SaaS lands
6
2
15
424
Just goes to show that one silly reply from me will take you over the 500k impressions. What are you waiting for? Let’s #connect
11
17
371