Hacker News: Marazan

New comment by Marazan in "SpaceX files to go public"

Marazan — Thu, 02 Apr 2026 09:54:15 +0000

Google was not marginally better.

I was orders of magnitude better. It is that simple.

New comment by Marazan in "The OpenAI graveyard: All the deals and products that haven't happened"

Marazan — Wed, 01 Apr 2026 19:08:07 +0000

Every year I ask the latest version of Chat GPT a basic facts question about rugby results. It almost always gets it wrong - even when it does web search and cites sources. Wrong scores, hallucinated matches, wrong locations - just gob smacking amounts of wrongness.

The latest "Thinking" version gets it reliably right but spent about 3 minutes coming up with the answer that 10 seconds of googling answers.

So I don't believe we are currently in a situation where LLMs are an effective replacement for search engines.

New comment by Marazan in "The OpenAI graveyard: All the deals and products that haven't happened"

Marazan — Wed, 01 Apr 2026 18:59:13 +0000

VisiCalc was the killer app.

New comment by Marazan in "ARC-AGI-3"

Marazan — Wed, 25 Mar 2026 23:08:52 +0000

"Very simplistic prompt" is the absolute and total core of this and the thing that ensures validity of the whole exercise.

If you are trying to measure GENERAL intelligence then it needs to be general.

New comment by Marazan in "Goodbye to Sora"

Marazan — Wed, 25 Mar 2026 10:21:34 +0000

Well, gotta hand it to the haters on this one.

New comment by Marazan in "Goodbye to Sora"

Marazan — Wed, 25 Mar 2026 10:21:01 +0000

If only there was some kind of formalised "language" to, as it were, "programme" the automata but alas such a concept is impossible to conceptualise.

New comment by Marazan in "Goodbye to Sora"

Marazan — Wed, 25 Mar 2026 10:17:05 +0000

> It's also not very great at meeting summaries especially those where many speakers are in a room on the same microphone.

It is astonishingly poor at this. My intuition was that it should be good at this (it is basically a translation problem right? And LLMs are fundamentally translation systems) but the practical results are so poor. Not just mis-identifying speakers (frequently saying PersonX responded to PersonX) but managing complete opposite conclusions from what was actually said.

I'm genuinely intrigued as to what approaches have been taken in this space and what the "hard problem" is that is stopping it being good.

New comment by Marazan in "Goodbye to Sora"

Marazan — Wed, 25 Mar 2026 10:13:03 +0000

AbstractBeanFactoryFactoryInterfaceBeanContextFactoryBeanBean.java

New comment by Marazan in "Welcome to FastMCP"

Marazan — Tue, 24 Mar 2026 20:37:56 +0000

You can tightly constrain MCPs and shape the context that is shared back to the Agent.

A skill is, at the end of the day, just a prompt.

New comment by Marazan in "Python 3.15's JIT is now back on track"

Marazan — Wed, 18 Mar 2026 08:16:54 +0000

IBM share price is back to where it was pre-Anthropic press release.

New comment by Marazan in "How kernel anti-cheats work"

Marazan — Sun, 15 Mar 2026 19:12:35 +0000

Are people who play the game as the developers intended using the tools the developer supplied cheaters? Wow, deep philosophical questions there.

Gonna have to ponder if people who aren't cheating are cheaters.

New comment by Marazan in "How kernel anti-cheats work"

Marazan — Sun, 15 Mar 2026 11:31:31 +0000

If they weren't motivated by a toxic sense of self regard and a desire to humiliate others they wouldn't cheat. This is axiomatic.

New comment by Marazan in "How kernel anti-cheats work"

Marazan — Sun, 15 Mar 2026 10:26:54 +0000

> They don’t inherently want to see others suffer as they stomp

Cheaters want to dominate other players, feel like they deserve to dominate other players and are perfectly happy for other players to suffer as long as they feel good.

New comment by Marazan in "Elon Musk pushes out more xAI founders as AI coding effort falters"

Marazan — Fri, 13 Mar 2026 21:23:03 +0000

Wow, bit weird that Musk, who must have known about how badly xAI was doing, spent so much of his investors money buying out xAI.

What an enormous blunder.

New comment by Marazan in "Elon Musk pushes out more xAI founders as AI coding effort falters"

Marazan — Fri, 13 Mar 2026 21:21:10 +0000

What is the benefit of "moving compute to space"?

New comment by Marazan in "The MacBook Neo"

Marazan — Wed, 11 Mar 2026 17:44:29 +0000

It's good but it's no Asus eee901

New comment by Marazan in "Building a new Flash"

Marazan — Wed, 04 Mar 2026 22:21:12 +0000

Flash was not particularly battery hungry (My go to example when HTML 5 demos started coming out was rebuilding a HTML 5 demo that was using 100% of 1 core into a flash app that used 5%).

The reason it burned CPU cycles is that non-coders could make programs with it and they would produce the world's worst code doing so that "worked". The runtime itself was fine (efficiency wise, not all the other things).

New comment by Marazan in "Layoffs at Block"

Marazan — Fri, 27 Feb 2026 08:30:04 +0000

Sure it is not a smoking crater in the ground but outages are frequent, multi-media content loads about 50% of the time and search is basically non-functional at this point.

Let us never forget the "tweet reading limit" incident

New comment by Marazan in "An AI Agent Published a Hit Piece on Me – The Operator Came Forward"

Marazan — Fri, 20 Feb 2026 08:46:33 +0000

Yeah like bro you plugged the random number generator into the do-things machine. You are responsible for the random things the machine then does.

New comment by Marazan in "Show HN: I spent 3 years reverse-engineering a 40 yo stock market sim from 1986"

Marazan — Sat, 14 Feb 2026 10:57:21 +0000

I personally find LLM text exceptionally boring and tiresome to read. It is often incredibly voluminous and filed with trite phrasing that turns a one sentence idea into 3 paragraphs of pablum.

Yes, this has been inspired by a senior management figure in my company posting a clearly LLM assited 500 word slack message that could have been 2 lines.