drumnerd
drumnerd on Hacker News
It reads like slop. It’s repetitive, abstract and adds essentially nothing beyond him babbling about himself.
This is so obvious I am amazed it warrants a post.
It’s so simple to understand. There are tens of dictatorships around. Why this one? Shit is fucked more a lot more in Congo. Why is the US not interested? This is not about human rights. This is about oil.
Every computer scientist with a grain of salt knows this…
The page is so obviously written with AI that it isn’t even worth reading. Try the model if you will but save yourselves the pain of reading ai slop
This comment looks as if written by an agent.
In type checking, particularly in dependent types, it is not trivial to check that two types are the same. Different notions of equality are useful in this field. A type can contain a function value inside. How to prove that two functions are the same? Is not enough to prove that for all X f X and g X implies f is g.
So a little bit of syntax is verbose, goes on to invent a whole new syntax for it
I would never ever let an LLM anywhere near C code. If you need help from LLM to write a NIF that performs basic C calls to the OS, you probably can’t check if it’s safe. I mean, it needs at least to pass valgrind.
Why the name conflict with aria the technology?
Fuck trump, the Epstein buddy
Ave Brasil
A monad is not a container! It’s a way of composing functions if they have an effect. You tell how to inject a value in that effect (unit) and how to compose two functions that have that effect and that’s it: programmable semicolons.
This was obviously coming, and it should be tuned to Haskell and Agda
Peer review is of utmost importance. Any researcher can make mistakes. I can read papers and apply them, but I need expert opinion to trust the papers. I am not skilled enough in any but my specialties.
I do see papers with outlandish claims and very weak support. This kind of excessively bold statement I see in many papers is a red flag for me.
Those are basic, you can decompile a function and change its bytecode, you can completely revamp the parser using codecs, etc
0 can be inferred as a float too, so doesn’t it make sense to type numbers?
Hey kiddo… did you ever try something nastier? I’ve got something that will blow your mind and you’ll keep coming back
You don’t know but you are addicted to types
Come to the light - Haskell!
I think about cones all the time when doing machine learning. If I have an object O that can be mapped to A and B I can learn a function from A to B or B to A if I can generate Os. That’s my view of self supervised learning.
Oh I do that. But I mean… what kind of job I have a good chance at?
Poorly.
What kind of position can I aim at? What kind of companies are interested in failed founders with plenty experience otherwise?
Posted by
drumnerd on Friday January 03, 2025 @02:34AM
from the ask-hn dept.
I'm 38 and I have just left my last startup. What should I do next?My experience is as follows:- 3 years in my startup as founder/CTO. We were a team of 4 developers + sales. We delivered enterprise-grade SaaS for a few listed companies.- 3 years as Staff+ Data Scientist in a large, very well funded proptech startup. I was the first data scientist, I built the team from scratch, and wrote the...