henriquegodoy

6 ms·

1.
Why models write slop: the environments are too small(henriquegodoy.com)
2 points by henriquegodoy 3mo ago|0 comments
2.
Extract-0: A specialized language model for document information extraction(arxiv.org)
195 points by henriquegodoy 1y ago|58 comments
3.
Learning Perl in one day and the importance of building strong foundations(guilhermenl.dev)
3 points by henriquegodoy 1y ago|0 comments
4.
Alvorada-Bench: Can Language Models Solve Brazilian University Entrance Exams?(arxiv.org)
1 points by henriquegodoy 1y ago|0 comments
5.
by henriquegodoy 1y ago
This is actually really needed, current ai design tools are so predictable and formulaic, like every output feels like the same purple gradients with rounded corners and that one specific sans serif font that every model seems obsessed with
6.
by henriquegodoy 1y ago
This is pretty cool and feels like we're heading in the right direction, the whole idea of being able to hop between devices while claude code is thinking through problems is neat, but honestly what excites me more is the broader patte
7.
by henriquegodoy 1y ago
Looking at this evaluation it's pretty fascinating how badly these models perform even on decades old games that almost certainly have walkthroughs scattered all over their training data. Like, you'd think they'd at least bru
8.
by henriquegodoy 1y ago
Thats incredible to see how ai models are improving, i'm really happy with this news. (imo it's more impactful than the release of gpt5) now, we need more tokens per second, and then the self-improvement of the model will accelera
9.
by henriquegodoy 1y ago
That SWE-bench chart with the mismatched bars (52.8% somehow appearing larger than 69.1%) was emblematic of the entire presentation - rushed and underwhelming. It's the kind of error that would get flagged in any internal review, yet h
10.
by henriquegodoy 1y ago
I dont think there's so much difference from opus 4.1 and gpt-5, probably just the context size, waiting for the gemini 3.0
11.
by henriquegodoy 1y ago
I think this blog post was the best way to get into Anthropic, and it was well-deserved. That's the reality of hiring in tech: there are many non-technical people judging whether technical people are competent or not. Escaping that mat
12.
by henriquegodoy 1y ago
Seeing a 20B model competing with o3's performance is mind blowing like just a year ago, most of us would've called this impossible - not just the intelligence leap, but getting this level of capability in such a compact size. I t
13.
by henriquegodoy 1y ago
I'm seeing a real-world example of Jevons paradox playing out here. When AI coding tools first emerged, everyone predicted mass developer unemployment. Instead, I'm watching demand for skilled developers actually increase. What&#x
14.
by henriquegodoy 1y ago
Ever thinked on automating this process of creating this side projects? i think that more and more future feels like a lot of ones having really big swarms of "agents" that can like research about ideas on the internet (like findi
15.
by henriquegodoy 1y ago
my vision is that the market is not really prepared for that right now, the best way is this guys is solving a really niche problem with their plataform and then expanding trough more areas
16.
by henriquegodoy 1y ago
Nice, i think that yall are on the correct path betting on evals, but please make your ui less "generic"
17.
by henriquegodoy 1y ago
Will apply this for the next interfaces that im going to build
18.
by henriquegodoy 1y ago
can i automate the process of answering this pr questions too?
19.
by henriquegodoy 1y ago
It's cool to see the perspective that many problems (somekinda communication problems, look at lawyers, compliance and etc...) can be solved by treating AI less as agents and more as modular components within a larger system. Once we b
20.
by henriquegodoy 1y ago
The point is that you can have a highly advanced teacher with infinite patience, available 24/7—even when you have a question at 3 a.m is game changer and people that know how to use that will have a extremaly leverage in their life.
21.
by henriquegodoy 1y ago
I've been tinkering with agentic systems for a while now, and this post nails some key pain points that hit close to home. The emphasis on splitting context and designing tight feedback loops feels spot on—I've seen agents go off
22.
by henriquegodoy 1y ago
crazy how the behaviour impacts the language and language impacts the behaviour such like a loop
23.
by henriquegodoy 1y ago
Speaking of interfaces, when will we have one that works just by thinking—something less intrusive than Neuralink—that lets us control not just Blender, but the entire computer? I think my productivity would increase a lot...
24.
by henriquegodoy 1y ago
IMO AI as a whole is just the catalyst
25.
by henriquegodoy 1y ago
Great post! i've been thinking along similar lines about human-AI interfaces beyond the copilot paradigm. I see two major patterns emerging: Orchestration platforms - Evolution of tools like n8n/Make into cybernetic process design
26.
Lessons I'd Tell My 12-Year-Old Self(henriquegodoy.com)
2 points by henriquegodoy 1y ago|0 comments
27.
Show HN: Interactive Enigma Machine Simulator(enigmasimulator.com)
12 points by henriquegodoy 1y ago|0 comments
28.
Show HN: FSSG – A Distraction-Free Writing Space with Focus Tracking(future-seems-so-good.com)
2 points by henriquegodoy 2y ago|0 comments
29.
by henriquegodoy 2y ago
Location: Brazil (Available 4AM to 9PM EST)/Remote Remote: Yes Wiling to realocate: yes Résumé/CV: https://drive.google.com/file/d/1VbOYflwT3crNqzCq8rQH7Mz31dC... / https://www.linkedin.c