Hacker News: josu

New comment by josu in "LLM Usage in Debian: Three Proposals"

josu — Sun, 26 Jul 2026 21:38:13 +0000

I was responding to this "merely produces syntactically likely combinations of the training data".

New comment by josu in "LLM Usage in Debian: Three Proposals"

josu — Sun, 26 Jul 2026 09:46:11 +0000

Yes, like the Jacobian counterexample.

New comment by josu in "LLM Usage in Debian: Three Proposals"

josu — Sun, 26 Jul 2026 06:33:09 +0000

The combination of the most probable tokens doesn't necessarily have to be in the training set, thus creating something completely novel.

New comment by josu in "Cloudflare Drop"

josu — Thu, 09 Jul 2026 10:50:21 +0000

It's a spectrum. I don't think that they will be able to sell t-shirts using a drawing you've uploaded. But it will probably allow them to defend themselves a bit better if they get sued for selling the data for LLM training.

New comment by josu in "Nintendo announces new product revisions in Europe with replaceable batteries"

josu — Mon, 06 Jul 2026 18:01:22 +0000

The console one seems the only relevant one. I'm a casual gamer, but the joycon running out of battery doesn't feel that annoying to me. That 16% has a much smaller impact on the JoyCon than on the main device.

New comment by josu in "Nintendo announces new product revisions in Europe with replaceable batteries"

josu — Mon, 06 Jul 2026 15:55:35 +0000

Is it?

>Battery capacity: 5172mAh, approximately 1% smaller than current version (5220mAh)

New comment by josu in "Spain Orders Blacklist of Palantir from Public and Private Companies"

josu — Thu, 02 Jul 2026 20:21:09 +0000

>Spain is not Somalia, why not let Indra do it?

The data may be safer with the CCP, at least they won't lose it.

Show HN: HN Who Is Hiring trends by month and category

josu — Tue, 23 Jun 2026 18:28:36 +0000

Article URL: https://hnhiring.josusanmartin.com/

Comments URL: https://news.ycombinator.com/item?id=48649249

Points: 2

# Comments: 1

New comment by josu in "CPanel's Black Week: 3 New Vulnerabilities Patched After Attack on 44k Servers"

josu — Sat, 09 May 2026 21:57:50 +0000

So CPanel's security is just as bad as their UI, who would have thought?

New comment by josu in "A recent experience with ChatGPT 5.5 Pro"

josu — Sat, 09 May 2026 15:23:18 +0000

LLMs can also be really good in fields where you are not an expert. You just need to be very aware of your limitations, and start parallel conversation so one agent fact checks the other.

New comment by josu in "Gemini Robotics-ER 1.6"

josu — Wed, 15 Apr 2026 17:05:53 +0000

Done! The whole planet is now veggies.

New comment by josu in "Spain to expand internet blocks to tennis, golf, movies broadcasting times"

josu — Tue, 14 Apr 2026 19:53:58 +0000

>enact similarly stupid laws.

No new law was enacted. The ISPs are enforcing a court order.

New comment by josu in "Waymo seeking about $16B near $110B valuation"

josu — Mon, 02 Feb 2026 16:46:03 +0000

>I wonder what will happen to the drivers if a large representation of the 1 million+ daily trips are displaced by automation?

If it happens gradually enough, they will just find other jobs. After the transition, society will be producing more with the same labor force, and thus the aggregate utility will increase.

New comment by josu in "SoftBank in talks to invest up to $30B more in OpenAI"

josu — Wed, 28 Jan 2026 14:59:43 +0000

> invested in (...) anything else practical.

I don't understand how this is the top comment. LLMs have unlocked a lot of value for me personally, and arguably for the society as a whole. They are also one of the coolest technologies I've tried in years. As a technologist, I'm really glad that money is pouring in and allowing us to find its limits.

New comment by josu in "I vibecoded my way into the #1 position on the Highload.fun leaderboard"

josu — Sun, 18 Jan 2026 21:44:35 +0000

How is it cheating? Do you also need to independently discover Bloom filters to be able to use them?

I vibecoded my way into the #1 position on the Highload.fun leaderboard

josu — Sun, 18 Jan 2026 21:04:00 +0000

Article URL: https://josusanmartin.com/blog/2026/01/18/the-game-has-changed-vibecoded-highload.html

Comments URL: https://news.ycombinator.com/item?id=46672143

Points: 4

# Comments: 2

New comment by josu in "Opus 4.5 is not the normal AI agent experience that I have had thus far"

josu — Tue, 13 Jan 2026 22:12:36 +0000

I think you misunderstood, it's not about solving the problem, is about finding the most efficient solution. Give it a shot, and see if you can get to the top 10 on any task.

New comment by josu in "Opus 4.5 is not the normal AI agent experience that I have had thus far"

josu — Fri, 09 Jan 2026 14:29:12 +0000

Thank you. Yeah, I'm doing all those things, which do get you close to the top. The rest of things I'm doing are mostly micro-optimizations such as finding a way to avoid AVX→SSE transition penalty (1-2% improvement).

But I don't want to spoil the fun. The agents are really good at searching the web now, so posting the tricks here is basically breaking the challenge.

For example, chatGPT was able to find Matt's blog post regarding Task 1, and that's what gave me the largest jump: https://blog.mattstuchlik.com/2024/07/12/summing-integers-fa...

Interestingly, it seems that Matt's post is not on the training data of any of the major LLMs.

New comment by josu in "Opus 4.5 is not the normal AI agent experience that I have had thus far"

josu — Wed, 07 Jan 2026 19:15:54 +0000

> The above is a software engineering problem. Reimplementing a JSON parser using Opus is not fun nor useful, so that should not be used as a metric.

I've also built a bitorrent implementation from the specs in rust where I'm keeping the binary under 1MB. It supports all active and accepted BEPs: https://www.bittorrent.org/beps/bep_0000.html

Again, I literally don't know how to write a hello world in rust.

I also vibe coded a trading system that is connected to 6 trading venues. This was a fun weekend project but it ended up making +20k of pure arbitrage with just 10k of working capital. I'm not sure this proves my point, because while I don't consider myself a programmer, I did use Python, a language that I'm somewhat familiar with.

So yeah, I get what you are saying, but I don't agree. I used highload as an example, because it is an objective way of showing that a combination of LLM/agents with some guidance (from someone with no prior experience in this type of high performing architecture) was able to beat all human software developers that have taken these challenges.

New comment by josu in "Opus 4.5 is not the normal AI agent experience that I have had thus far"

josu — Wed, 07 Jan 2026 19:06:13 +0000

You are looking at this wrong. Creating a json parser is trivial. The thing is that my one-shot attempt was 10x slower than my final solution.

Creating a parser for this challenge that is 10x more efficient than a simple approach does require deep understanding of what you are doing. It requires optimizing the hot loop (among other things) that 90-95% of software developers wouldn't know how to do. It requires deep understanding of the AVX2 architecture.

Here you can read more about these challenges: https://blog.mattstuchlik.com/2024/07/12/summing-integers-fa...