Kacper Kazimierak (Casper)

I am closing in on 100 followers, and I am extremely happy and grateful that so many of you decided it's worth following me. I started regularly using X 1,5 months ago, and I still remember my joy from the first subscriber, then 10, and it's still growing. The growth is not as fast as for some other accounts on X around the same time, but I decided to follow a few simple rules to provide quality content: - no copy pasting other posts - this is annoying when you see the same content copied and pasted between different accounts, that's why Quote and repost functions exist - I never post AI generated content - I occasionally use Grammarly for typos and grammar checking, and that's it. I believe people are not here for generated content, but the real human opinions and stories - I try to add something meaningful to the conversations instead of vague responses or even worse, “let's connect” spam And recently I noticed it is starting to show some promising results. I regularly post my comments on tech, software engineering, and AI from a grounded perspective. I especially found it fun and valuable to debunk AI hype/fearmongering and disinformation. I also share updates on my projects and stories from my career that I find interesting. I use my decade of experience as a software engineer, along with common sense and critical thinking, to shape my perspective. If this sounds interesting - follow for more content.

@jailbreakersAI @UnathiAfrika @LuizaJarovsky that's right

@jailbreakersAI @UnathiAfrika @LuizaJarovsky If you use it as a tool then yes, I agree. The problem is that many users are treating it as a “companion” which AI is definitely not. And LLM providers are encouraging such approach because it benefits them.

Nobody wants to read AI-generated posts. Nobody wants to read AI-generated emails. Nobody wants to read AI-generated newsletters. Nobody wants to read AI-generated reports. Nobody wants to read AI-generated articles. Nobody wants to read AI-generated anything.

@UnathiAfrika @LuizaJarovsky well, that would be the best solution, right now everyone hopped on AI train, without really thinking much about consequences of using it in a wrong way

@TheByteRacoon @LuizaJarovsky I agree. So I think the challenge is to train people to know how to use it correctly, thing let it replace their thinking and accept its output as gospel truth. Banning it or rejecting its output outright will not help either.

@UnathiAfrika @LuizaJarovsky As long as you treat AI as a tool, I think there is nothing wrong with that. The problem is that some people treat it as a person and accept AI output without critical thinking.

@TheByteRacoon @LuizaJarovsky Perhaps learn to work WITH AI, not let AI work FOR you. That said, I wonder if you'd say the same with AI applications in Robotics.

it's not even close, you are comparing apples to potatoes. AI generated content on the other hand, is its interpretation of the topic, it's not deterministic, not flawless, and is trying to take away critical thinking from the brain. Calculators are generating fully deterministic results, always the same, it's just a tool to reduce load on the brain. I see a huge gap between those two examples, but maybe your ability to think critically is already atrophied.

@LuizaJarovsky Sometime in the 1960s 😊😊😊

@atmoio I wonder how long Altman is going to stay afloat with such brilliant ideas. It's not like the investors are going to forget about their billions… right? They are in panic mode after realising transformers are maybe not the viable path to AGI

Sam Altman is giving up on cancer and pivoting to erotica.

@sdrzn so the people who will retain their ability to code, or even better - to think without an AI will be the winners in a few years? 🤔

every major ai lab is losing money on you right now. openai lost $5B in 2024 on $3.7B revenue. they burned ~$8B on inference alone in the first 3 quarters of 2025. and they're forecasted to burn $218B in cash from 2026 to 2029. For comparison, uber burned $18.2 billion over six years before turning profitable. the strategy is subsidize inference to near-zero, get every engineering team dependent on their models, let switching costs build up quietly, then close the subsidy window once you're locked in. here's the part people get wrong. "but inference costs are falling! tokens are cheaper than ever!" yes, per-token costs dropped ~10x year over year. but that's for them, not for you. these companies have tens of billions in losses to recoup. cheaper infrastructure doesn't mean cheaper pricing when you're $218B in the hole. and as coding agents become the default way software gets built, the volume of tokens per developer is exploding. agentic loops hit the model 10-20x per task. context windows keep growing. your ai bill scales with every user, every feature, every agent running in the background. the unit cost goes down but the total spend goes up - that's the trap. this is why we built cline to be model-agnostic, because vendor lock-in to a single inference provider is dangerous. you should be able to swap models, run open source on your own infra, use whatever provider gives you the best price-performance for your workload - and never worry about a pricing rug pull.

@vivoplt battery, M chips performance almost without heating up, native unix, ecosytem if you have other apple devices, reliability

Honestly, why do developers choose MacBook over other laptops?

@ChShersh maybe on X, the real industry has a lot of inertia. I agree the engineering will look completely different, but IMHO certainly not in 60 days. Maybe 1-2 years.

I’ve just met some engineers doing incredible things with AI. The landscape of programming is changing so rapidly. My prediction: AI won’t replace engineers. But in 60 days engineering will look completely different.

@SamuelBeek I get it, but why Autodesk?

If you understand this meme, we friends

@adxtyahq ai or not, most of the devs I know now rather use Tailwind, Bootstrap or other css framework instead of writing from scratch. It cuts amount of CSS reuired by orders of magnitude

i’ve never seen anyone write raw css anymore did everyone just switch to ai?

@pcshipp your free tier seems to be to generous

How do people make $10K MRR or even $100K MRR while I’m still stuck at $9 MRR? - $9 MRR - $2 revenue - 198 New users - 206 Active users No idea how to start marketing

one year has finally passed... feels soo good

@md_kasif_uddin I'm an SWE at a Europe's large marketplace platform, during pandemic they did switch to a lot of remote employment

@TheByteRacoon Where do you work?

As a developer, which one do you prefer more? (Be honest) 👀

@sysxplore I'll take both and use WSL2 on Win 😅 Ok I'm currently on Mac, but before this seemed like the most optimal setup

@thekitze I have simple rule, most posts starting with "XXXX is dead" or "Holy shit xxxx just...." if it's not a sarcastic one = likely block

blocking is harsh so i have 1 rule about it: i block you if you insult me, period i don't block people about opinions so color me fkn surprised when i stumble on an anon i haven't directly interacted with and i see they've blocked me some bros just cannot stand opinions 💀

@thekitze fr it has a junior's tendency to create big messy, do-all functions

@TheByteRacoon ooooh!!

llms are such bullshitters it's incredible.. we setup an oxlint rule the other day for max 200 lines per file and it "magically disappeared" LOOK AT THIS MESS (btw max 200 lines per file is probably one of the best rules for llms to not write slop)

I don't want to defend meta, but this is mostly misinformation: - they are not shutting it down, just removed from Quest VR - 80bn is the budget for the whole Reality Labs budget, which was split among other (also hardware) projects, not Metaverse app. But yeah, when they rebranded Fb to Meta, and announced investment into Metaverse, I thought either I live in a bubble and this is some new trend that I missed, or they are delusional. Now at least I know it's not me who lives in a bubble 🤣

@hiarun02 simple projects that do not provide any value beside parsing some irrelevant data, like counters, csv parsers, etc. Instead of deploying end-to-end something actually working and accessible via web / package manager

What’s the most outdated thing still taught in computer science degrees?