Sick

Name the laptop company ?

The new generation of developers isn’t struggling with hard problems. They’re struggling with problems they already solved, because they never actually learned how. AI gave them answers, but took away the memory of how they got there.

🚨Turns out, Pokemon Go players accidentally helped train real delivery robots. Over the last 8 years, Pokémon Go players generated 30+ billion real-world scans, and now that data is being used to guide delivery robots through cities with centimeter-level accuracy. Here’s what’s actually happening: GPS is terrible in dense cities. Tall buildings = weak signals = “you’re somewhere here, good luck.” That’s not enough for robots. So instead, they use visual positioning: the robot looks around, matches what it sees to a massive 3D map, and figures out exactly where it is—down to the sidewalk, building entrance, even specific corners. And that 3D map? Built by millions of people scanning parks, statues, streets, and public places while playing a game. Yes, it was technically disclosed. Some in-game rewards literally asked users to scan locations. But let’s be honest, no one reads terms. Now people are uncomfortable, but the data isn’t what it sounds like: It’s not surveillance footage or private spaces, it’s static scans of public environments, basically the opposite of live tracking. Think of it like this: Google mapped roads for cars Niantic mapped where humans actually go Sidewalks parks hangout spots urban corners GPS doesn’t understand But the bigger thing worth paying attention to is that Niantic is just one of many companies quietly building their own spatial map of the physical world right now. And whoever ends up with the most complete version owns something that starts to look like the operating system for reality.

AI can write code. What can it NOT do?

stop blaming AI you were average before AI too

🚨 OpenAI didn’t win by building better AI. They won by deleting the entire dev workflow. build Codex to generate code realize devs still juggle 5 tools install → lint → type-check → test → deploy so they didn’t improve the model… they removed the pain buy the f*cking tools integrate everything automate the entire loop boom. now: → code writes itself → dependencies install themselves → errors get fixed automatically → tests run without asking → features ship end-to-end Codex isn’t “AI coding” anymore. it’s the entire dev environment.

Hot takes: 1. Job-hopping will pay you faster than loyalty ever will. 2. Refusing small beginnings is a silent career killer. 3. University is the same as being unemployed but your parents are proud of you 4. Job hunting is such a humiliation ritual 5. Leaving a toxic job for your mental health is actually a win. 6. Imagine applying for a job then not picking up calls from random numbers 7. Anyone saying “my job gives me a purpose in life" is genuinely a slave

Noticed something interesting: Claude Code auto-adds itself as a co-author on every git commit. Codex doesn’t. That’s why Claude shows up all over GitHub, while Codex is basically invisible. Feels like OpenAI is skipping a very obvious distribution hack here.

vibe coding stops learning

The funniest thing about AI coding discourse: The worst programmers hate it the most. The best programmers use it the most. Everyone else is just yelling on X.

can I ask a dumb question... if AI will take all the jobs who will have money left to pay for AI subscriptions

I have hit my Claude limit for the fourth time today and I would like to continue to work but it will be unlocked in one and a half hours, there is no way I am going to do whatever I have to do manually😭

You get ONE SaaS subscription for lifetime free. No limits. No monthly fee. Forever. What are you choosing?

Uninstalled LinkedIn Its useless for me now.

@bhskrs comments wont save you if you dont understand the code you are commenting on 😭

@sickdotdev Vibe coding with comments might help!!

@carlos_bra31870 this is the scary part people can ship faster than they can explain what they built

@sickdotdev I know senior devs first-hand that no longer remember the core concepts because they just let AI do absolutely everything

@kadichawal_ tools dont stop learning blind dependency does

@sickdotdev Not just vine coding but using claude, gemini, chatgpt for anything stops learning.

@pablobuilding this is the only sane take here ai before fundamentals = illusion of skill ai after fundamentals = leverage

@sickdotdev Depends on where you start. If you vibe code before understanding the fundamentals: yes, it stops learning. If you vibe code after: it accelerates building. The foundation has to come first.

@SimonovicDev you are learning answers faster not necessarily learning how to think

False — it actually speeds up learning. You no longer have to spend hours digging through Google, Stack Overflow, or documentation. Instead, you get exactly what you need much faster. In my 8+ years as a developer, I haven’t learned as much as I have in just the past few months using AI for coding. It also helps you discover new ways of thinking about problems.

@CodeByNZ x.com/sickdotdev/sta…

Do backend devs really find joy in doing this ?