amunozo's comments

On: A Replacement for BERT

[amunozo]

Yes, the paper says that is only English.

On: No More Adam: Learning Rate Scaling at Initializat...

[amunozo]

It's time to stop the "All You Need" titles. This one does not even sound good .

[v3ss0n → amunozo]

Need to write an article `All you need should consider harmful` .

[scotty79 → v3ss0n]

"Considered harmful articles are all you need"

[0xdeadbeefbabe → scotty79]

Goto is all you need.

[cuuupid → amunozo]

It's one of the most irritating snowclones because most of the time the papers are not presenting some dramatic leap forward like attention.

On: Veo 2: Our video generation model

[amunozo]

And places with these systems are those that achieved the best quality of life and peace.

On: Cheaper to rent in Barcelona and commute to London...

[amunozo]

An apartment like that could cost 1500€ today. Also, flights are also more expensive.

On: How I write code using Cursor

[amunozo]

I've been using Copilot in VSCode as I have it free as a student. I wanted to try Cursor, but money is tight. Are they that much different? If so, what makes Cursor so special?

On: Ask HN: Is there a market for LLM-powered local pr...

[amunozo]

I thought about that too! In the end, I'll probably do them both anyway.

On: Ask HN: Is there a market for LLM-powered local pr...

[amunozo]

What about consumer applications where privacy is not that much needed? What I mean is if there is any demmand for a traditional, pay-once software.

[haebom → amunozo]

I don't think it's worth it for personal use, and the reason is simple. There are already a lot of tools that can write local LLMs, so I don't think I'd be able to get users to pay for just using the model, I'd have to provide functionality, like automating certain behaviors.

[JohnFen → amunozo]

The only way I'd consider using any AI-involved system is if it's all local. SaaS is a dealbreaker for me on this.

I don't think that helps you, though, since (like most of the people on HN) I am not representative of the general public.

On: Ask HN: What's the most creative 'useless' program...

[amunozo]

Your webpage is gorgeous, congrats!

On: Math is still catching up to the genius of Ramanuj...

[amunozo]

Exposing students to other subjects is what make them discover what they like. I think the optimal balance would be studying a wide number of subjects, but being able to select for an even larger set.

On: Grandmaster-level chess without search

[amunozo]

It's a knowledge distillation. You can then use this smaller, more efficient models instead of the larger one.

[chvid → amunozo]

Or maybe it is just memorizing a very large number of games.

[azakai → chvid]

They address the possibility of memorization in the PDF:

> This effect cannot be explained by memorization since < 1.41% of the initial puzzle board states appear in our training set.

[tech_ken → chvid]

Seems more like a 'compression' of the large number of games, or even like an approximate 'index' of the database

[alkonaut → amunozo]

Is this network smaller than stockfish and by what metric is that?