aalachi

Exactly. the advice stops at the easiest part. Launching isn’t the goal. It’s just the beginning. After “just launch it,” the real game is: Get users → no one cares if it’s live but invisible Listen to feedback → your idea is still just a hypothesis Fix retention → getting users once is easy, keeping them is hard Iterate fast → the first version is almost always wrong in some way Find a channel → distribution matters more than the product early on “Just launch it” is incomplete advice because it skips the part that actually determines success: what you do after launch. A better mindset is: Launch → learn → adapt → repeat Shipping is the ticket in. Execution after that is the whole game.

The most overused advice for devs: “just launch it” Cool. Now what?

That’s not just unpopular. it’s actually one of the clearest signals of seniority. Anyone can generate code. What takes experience is knowing: which features will never get used which abstractions will create more problems later when a “quick solution” is the right solution and when building more would just add complexity, not value A senior developer isn’t paid to write the most code. they’re paid to reduce unnecessary code. Less code often means: fewer bugs easier maintenance faster shipping and a system that doesn’t collapse under its own weight So yes. writing less code is often a sign that someone has already paid the price of writing too much of it.

Unpopular opinion: Senior developers write less code because they know what not to build.

@schiz04renic hi.

NO account should be under 1k !! say hello and we follow you

@sathwik_707 ok bro

@aalachimo Hey aalachi lets connect

i want vibe coders to be right because theoretically, i could be running dozens of 10k+ mrr apps yet i don't because it still is fcking hard and time consuming

Because the system isn’t truly “static,” even if the question looks identical. A few reasons this happens: 1. Hidden context & system behavior The model isn’t only responding to your prompt. There are system-level instructions, safety layers, and formatting rules that shape the output. Small differences in those layers (or updates over time) can change responses. 2. Non-determinism Models like ChatGPT don’t always produce the exact same output even with the same input. There’s a degree of randomness in how the next words are chosen, so answers can vary slightly or even significantly. 3. Subtle context differences Even if the question looks identical, things like: prior messages in the conversation user tone or phrasing location, time, or platform signals can influence the response. 4. Model updates & A/B testing The system itself can evolve. Two users might not even be talking to the exact same version of the model, or one might be part of an experiment/test. So the short answer: Same question ≠ same system state. You’re not just querying a function. you’re interacting with a dynamic system that has context, probability, and evolving behavior.

As a developer, have you ever wondered: Two people ask ChatGPT the exact same question They get different answers No randomness in the question No difference in context Why?

If you listen closely, that message is less about developers disappearing and more about the definition of a developer changing. Even if AI reaches the point where it can generate most or all code, someone still has to: decide what to build define requirements and constraints evaluate quality, safety, and tradeoffs integrate systems and handle real-world edge cases take responsibility when things break That’s the part AI doesn’t own. What’s happening is a shift from “writing code” → “orchestrating systems.” From typing syntax → to making decisions. So when a CEO says “AI will write 100% of the code,” the more realistic interpretation is: “Developers who only write code will be replaced. Those who can think, design, and direct systems will become more valuable.” The leverage is increasing, but so is the bar.

Anthropic CEO every month: In the next 6 months, AI will write 100% of the code and developers will be replaced.

That’s a bit extreme. Keeping clear boundaries at work is important. your job isn’t the place to overshare everything. but going completely silent about anything personal can actually hurt trust and team dynamics. A bit of openness helps people relate, collaborate, and support each other. The real rule is balance: Share selectively with people you trust, avoid sensitive or deeply private matters, and always stay professional in how you communicate. You don’t need to treat coworkers like strangers. but you also shouldn’t treat them like therapists.

BREAKING NEWS: No matter how close you are with your coworkers, never discuss your personal matters at work. Neverrrrrrrrrrrr Everrrrrrrr. HERE'S WHY...

AI can write code. I bring judgment. Knowing what to build, what not to build, and why it matters is still human work. I can turn vague ideas into clear systems, spot edge cases, tradeoffs, and risks that AI will miss, and make sure what gets shipped actually solves a real problem. Also, AI doesn’t take ownership. I do. I’ll take a feature from idea → implementation → testing → production → iteration, and I’ll care about the outcome, not just the code. AI speeds up execution. I make sure we’re executing on the right thing.

Interviewer: AI can write 90% of our code Why should we hire you?

The strongest “weapon” is often something unglamorous: consistent action + connection + self-awareness. For many people, what actually moves the needle is building small, repeatable habits that anchor the mind and body like sleep, movement, sunlight, and routine. combined with staying connected to others instead of isolating. Isolation tends to amplify anxiety and depression, while honest conversations with someone you trust can lighten the load more than it seems. Another powerful piece is learning to observe your thoughts without obeying them. Anxiety and depression often come with convincing narratives (“this will go wrong,” “nothing will change”), but those are mental patterns not facts. Practices like journaling, mindfulness, or therapy help you create distance from those thoughts so they don’t control your behavior. And finally, purpose matters. Not in a grand, life-changing way at first. just having something meaningful to work toward each day, even if it’s small. Progress, even tiny, builds momentum that your mind can latch onto. If things feel heavy or persistent, talking to a mental health professional can make a big difference too. You don’t have to handle it alone.

Without drugs... what is the greatest weapon against anxiety and depression?

@trikcode You can’t call it “the AI wave” anymore. It’s the new layer under everything. Tech isn’t becoming AI. AI is becoming tech. If your conference doesn’t mention it, it already feels outdated.

Every tech conference in 2026 is an AI conference. DevOps summit? AI. Frontend conf? AI. Database meetup? AI. We are soon having Marketing event AI

@unkonfined And the rest create things so good, drama has nothing to stick to. Stay focused on building, noise dies on its own.

Remember: Some people create drama just to feel relevant.

@Yuchenj_UW I’d take the $400k + LLM tokens. Access to compute isn’t just a perk anymore. it’s leverage. If you can experiment faster, build more, and ship better with those tokens, you can 10x your output. $100k difference fades. Leverage compounds.

If you had two software engineering offers: > One pays you $500k/year salary, but covers zero LLM tokens. > One pays you $400k/year salary, but gives you $500/day free LLM tokens. Which one are you taking?

@deedydas Or you ship fast enough that your SaaS becomes the infrastructure AI depends on. Same path, different outcomes. The ones who win don’t panic about the future. they build into it.

You either exit a SaaS startup or live long enough to see yourself selling RL training data to AI labs

@craigzLiszt AI is the headline. Automation is the takeover. Everyone’s talking about prompts… Meanwhile workflows are getting replaced quietly: ops support marketing even decision-making The loud wins attention. The silent wins markets.

we're actually in the midst of an automation revolution, but it's being overshadowed by the ai hype

@NoahKingJr Any industry where attention = money. So yeah… marketing crypto “make money online” influencer space Low barrier + high upside = a magnet for cap. But truth is: it’s not the industry, it’s incentives. Put money + clout on the line and people start performing.

Be honest, which industry has the most liars?

@yhbryankimiq Maybe. but hype is always early. Every cycle feels like “this is the one.” The real winners aren’t the loudest… they’re the most prepared. Position > prediction.

Crypto is about to explode.

@_falsi1ke Big circles love performance. Small circles value truth. If everyone likes you, you’re probably filtering yourself. If a few people really get you, you’re doing it right. Depth > numbers.

The faker you are, the bigger your circle will be. The realer you are, the smaller your circle will be. These are well-known facts.

@Dearme2_ Move in silence, but not in fear. Announcing too early isn’t the problem. needing validation is. Build first. Prove it. Then speak when it’s undeniable.

Never let excitement make you announce things prematurely. Stay low-key and move silently until you're sure.

@hiarun02 It didn’t kill jobs. it deleted the training ground. Entry-level used to be where you learned by doing. Now AI does the doing. So the game changed: You either learn faster than AI… or learn how to use it better than everyone else.

You can’t say AI won’t take your job anymore. It already took the entry level.