Marcin Doliwa

"Fuck your Show HN with 1 point and zero comments..." I built a site that generates a daily 25th Hour mirror rant about the Hacker News front page. Edward Norton vs. the tech industry, every single day :) AI turns out to be really good at being angry about tech. 👇👇👇

@DmytroKrasun Try something offline, should be easier.

I regret losing that skill of building stuff for pure fun. I feel guilty now if I don't work on anything that doesn't grow my MRR. Was super optimized and focused for too long. Don't know if I will recover.

I have the impression that Anthropic, Google, OpenAI just wait until someone releases a model, then they release their own right after. Google released Gemini 3 Pro, 1-2 weeks later we had Opus 4.5 I bet it was ready way before

@johnrushx I would be a lumberjack, and that would be ok.

What work would you do if money wasn't a concern?

@thdxr Warming up X algorithm before some announcement ? 😀

parenting tip kids can't really remember things until they're 3-4 so no point interacting with them until then

@rrmdp Do you have any examples?

So many job boards and startups making good $$$ without a .com domain. crazy how "I need a .com" is still used as an excuse to not launch ship the damn job site!!!

@antoniogm Had to check if it's a parody account or not.

Nothing better than the office on a Sunday: peace, solitude, focus, capitalism momentarily dormant.

@Dan_Jeffries1 @akshay_pachaar Why? With their resources it looks like a good idea to let others play with their findings and wait until someone figures out what to do with it then use their resources to become a major player.

@akshay_pachaar Cool paper but the reason we know this is not Attention is All You Need v2 is because Google will never ever ever ever ever let their researchers publish something as broad and impactful as AIAYN ever again. :) They only let them publish lesser papers now.

Google just dropped "Attention is all you need (V2)" This paper could solve AI's biggest problem: Catastrophic forgetting. When AI models learn something new, they tend to forget what they previously learned. Humans don't work this way, and now Google Research has a solution. Nested Learning. This is a new machine learning paradigm that treats models as a system of interconnected optimization problems running at different speeds - just like how our brain processes information. Here's why this matters: LLMs don't learn from experiences; they remain limited to what they learned during training. They can't learn or improve over time without losing previous knowledge. Nested Learning changes this by viewing the model's architecture and training algorithm as the same thing - just different "levels" of optimization. The paper introduces Hope, a proof-of-concept architecture that demonstrates this approach: ↳ Hope outperforms modern recurrent models on language modeling tasks ↳ It handles long-context memory better than state-of-the-art models ↳ It achieves this through "continuum memory systems" that update at different frequencies This is similar to how our brain manages short-term and long-term memory simultaneously. We might finally be closing the gap between AI and the human brain's ability to continually learn. I've shared link to the paper in the next tweet!

@csaba_kissi Do you have some kind list with all of them ?

At 33, I quit my job. Since then, I've launched over 200 side projects and earned multiple 7 figures 💰EUR online. ❌ No CS degree. ❌ No Bootcamp. ❌ No internship. ❌ No certificates. ✅ Just self-taught programming skills and English. Remember, nothing is impossible! Good luck to everybody!

Any theories on why joysticks usage died?

@marclou Who has the lowest ratio followers/revenue?

Startup revenue by founder’s 𝕏 follower count: - 0 - 1K followers: $864 - 1K - 10K followers: $5,268 - 10K - 100K followers: $100,373 - 100K - 1M followers: $750,337

@thoughtlesslabs Decades ago we did it with music albums 😀

Holding my computer like this at the library to impress people

@thdxr Hobbies are even dumber, you do it for free or very often lose money on it.

it's ok if your github is empty it doesn't matter but now there's idiots that are like "it's dumb to do work that you're not paid for" 1. no one wants your work free or paid 2. that's like saying everyone that does charity work is dumb

@TheAnkurTyagi Same as humans when they are not aware that they're lying.

AI, has one huge problem. It lies and it does it with supreme confidence.

@ThePrimeagen @BHolmesDev Is there any way you recommend to have similar experience in neovim ?

@BHolmesDev I love inline completion Cursor tab is literally one of the best things ever

To all the devs that either haven’t used agentic coding tools, or stopped using them after a while: what turns you off about them?

@ApoStructura @levelsio Lol, same here. I asked my wife to read it to make sure all is fine with me. 😀

@levelsio I had to read another post to convince myself I wasn’t having a stroke and I could still understand English

I wasn't a big fan of Zuck but he was of the once there and before that was definitely ever been Many people think he could not be the to be the of between but then he do. And not only they are but he was with Meta

I finished a feature working with @opencode and my last message was "ok, it works now", or other times when I want to quit I send "exit". I'm wondering how much energy/money is lost on these ending conversation messages :)

@benjamincrozat I was in a similar spot (in my case my "office" was a small desk in our bedroom) and moved to a co-working space, which I totally recommend.

Dad developers. How can you even work at home when you have a kid and a pregnant wife who are always interrupting you? Yes, I have my own office. But it seems like they are mistaking it for a hallway or something.

my 2 month old baby slept for 12 hours i guess i was right you all just suck