Hacker News: rocho

New comment by rocho in "The Webpage Has Instructions. The Agent Has Your Credentials"

rocho — Sun, 15 Mar 2026 18:13:19 +0000

I absolutely agree, although even that doesn't solve the root problem. The underlying LLM architecture is fundamentally insecure as it doesn't separate between instructions and pure content to read/operate on.

I wonder if it'd be possible to train an LLM with such architecture: one input for the instructions/conversation and one "data-only" input. Training would ensure that the latter isn't interpreted as instructions, although I'm not knowledgeable enough to understand if that's even theoretically possible: even if the inputs are initially separate, they eventually mix in the neural network. However, I imagine that training could be done with massive amounts of prompt injections in the "data-only" input to penalize execution of those instructions.

New comment by rocho in "Gemini 3.1 Pro"

rocho — Thu, 19 Feb 2026 20:32:48 +0000

I find Gemini is outstanding at reasoning (all topics) and architecture (software/system design). On the other hand, Gemini CLI sucks and so I end up using Claude Code and Codex CLI for agentic work.

However, I heavily use Gemini in my daily work and I think it has its own place. Ultimately, I don't see the point of choosing the one "best" model for everything, but I'd rather use what's best for any given task.

Engineers are becoming sorcerers – Future of software dev with OpenAI Sherwin Wu

rocho — Sun, 15 Feb 2026 07:53:28 +0000

Article URL: https://www.lennysnewsletter.com/p/engineers-are-becoming-sorcerers

Comments URL: https://news.ycombinator.com/item?id=47021857

Points: 1

# Comments: 0

Agent-native Architectures – A Technical Guide

rocho — Sat, 10 Jan 2026 06:29:11 +0000

Article URL: https://every.to/guides/agent-native

Comments URL: https://news.ycombinator.com/item?id=46563328

Points: 3

# Comments: 0

New comment by rocho in "Questions censored by DeepSeek"

rocho — Mon, 03 Feb 2025 14:11:10 +0000

That's not DeepSeek, it's a Qwen or Llama model distilled from DeepSeek. Not the same thing at all.

New comment by rocho in "Gukesh becomes the youngest chess world champion in history"

rocho — Fri, 13 Dec 2024 08:04:10 +0000

You don't need to propagate it, you just need to show the gradient of the current position alongside with the classical evaluation, to give more context to the viewers.

New comment by rocho in "OpenAI, Google and Anthropic are struggling to build more advanced AI"

rocho — Tue, 19 Nov 2024 10:07:12 +0000

That's correct. I saw a paper recently that showed how LLMs performance collapses when they are trained on synthetic data.

New comment by rocho in "How I write code using Cursor"

rocho — Sat, 02 Nov 2024 20:54:51 +0000

Why would they be? Cursor took an existing editor and added some AI features on top of it. Features that are enabled by a third party API with some good prompts, something easily replicable by any editor company. Current LLMs are a commodity.

New comment by rocho in "It's Time to Stop Taking Sam Altman at His Word"

rocho — Sat, 05 Oct 2024 20:16:22 +0000

To you maybe. But if Claude or any other competitor with similar features and performance keeps a lower price, most users will migrate there.

New comment by rocho in "It's Time to Stop Taking Sam Altman at His Word"

rocho — Sat, 05 Oct 2024 20:13:40 +0000

See "AI models collapse when trained on recursively generated data"

https://www.nature.com/articles/s41586-024-07566-y

New comment by rocho in "I Am Tired of AI"

rocho — Sat, 28 Sep 2024 19:52:08 +0000

But that does not prove anything. We don't know where we are on the AI-power scale currently. "Superintelligence", whatever that means, could be 1 year or 1000 years away at our current progress, and we wouldn't know until we reach it.

New comment by rocho in "Independent directors of 23andMe resign from board"

rocho — Wed, 18 Sep 2024 15:06:44 +0000

But a genetic analysis to find your "roots" is also something you do once. Does Ancestry have subscriptions for that?

New comment by rocho in "How many photons are received per bit transmitted from Voyager 1?"

rocho — Mon, 03 Jun 2024 20:36:05 +0000

Wow, I never thought about how Voyager communicates with Earth. But now I wonder: if Voyager just sends photons towards the Earth, at the receiving end how are we recognizing which photons are coming from Voyager and how is the "signal" decoded?

New comment by rocho in "The killer app of Gemini Pro 1.5 is using video as an input"

rocho — Thu, 22 Feb 2024 15:41:09 +0000

I'd consider installing it if it had:

* In-depth technical explanation with architecture diagrams

* Open-source and self-hosted version

Also I didn't understand if it talks to a remote server or not. Because that's a big blocker for me.

New comment by rocho in "23andMe changed its terms of service to prevent hacked customers from suing"

rocho — Wed, 13 Dec 2023 06:46:56 +0000

That's only true now. You don't know that DNA leakage won't be a higher risk in the future (and FWIW, my opinion is the opposite of yours regarding the future risks). Moreover you can change your passwords, but you can't change your DNA.

New comment by rocho in "23andMe changed its terms of service to prevent hacked customers from suing"

rocho — Wed, 13 Dec 2023 06:35:51 +0000

> Even my brother and I share only ~50% DNA.

This is completely false. Any two random humans have more than 99% overlap by virtue of being the same species. It's even higher for brothers. We also share around 90% DNA with cats, dogs and elephants.

https://www.amacad.org/publication/unequal-nature-geneticist...

> I'm not too worried about it because it's never a 100% overlap.

This doesn't make sense. If they were equal, you'd be the same person except for environmental differences. Many applications don't need equal DNAs. E.g.

https://youtube.com/watch?v=KT18KJouHWg

> About insurance companies, they're legally forbidden to use such data.

This is a very weak argument. There's a long history of companies doing illegal things, and even if it's illegal today it doesn't mean it'll be illegal tomorrow.

New comment by rocho in "Undermining Democracy: The EU Commission's Controversial Push for Surveillance"

rocho — Sun, 15 Oct 2023 08:32:42 +0000

They'd just bring it up shortly after under a different disguise.

New comment by rocho in "How (not) to apply for a software job"

rocho — Sat, 30 Sep 2023 12:34:50 +0000

Did you consider that your experience is but one of many? No need to be rude and bitter to the author.

New comment by rocho in "The boiling frog of digital freedom"

rocho — Sun, 03 Sep 2023 20:25:02 +0000

The options might exist, but that doesn't mean that they are viable/usable. I agree with many of your points, but regarding document processing, LibreOffice really does not compare with any of the best alternatives (Google Documents, Office Word). It lacks options, if the options exists they are hidden in obscure menus, it doesn't display many things correctly (at least on my Linux computer), etc. For document processing, I'm still stuck on Google Docs, which is quite a sad state of affairs.

New comment by rocho in "Blocked by Cloudflare"

rocho — Thu, 10 Aug 2023 08:58:06 +0000

With Firefox you can toggle some settings that will make much harder to generate useful fingerprints. That's already a massive privacy difference.