HN user

Al0neStar

492 karma

The quick brown fox jumps over the lazy dog.

Posts39
Comments70
View on HN
twitter.com 1y ago

Poh-Shen Loh: GPT-o1 got a perfect score on my CMU undergrad math exam

Al0neStar
4pts0
economictimes.indiatimes.com 1y ago

US Constitution disappears from White House website on day two of Trump 2.0

Al0neStar
6pts2
link.springer.com 1y ago

Paths: Why is life filled with so many detours? [Open Access]

Al0neStar
1pts0
www.sciencedirect.com 2y ago

Notes from Beethoven's Genome

Al0neStar
1pts0
blog.piekniewski.info 2y ago

Fat Tails Are Weird

Al0neStar
4pts1
twitter.com 2y ago

"30.1K BTC of Silk Road funds controlled by US gov is on the move "

Al0neStar
34pts10
www.sciencedirect.com 2y ago

Benchmarking LLMs against human expert-curated biomedical knowledge graphs

Al0neStar
41pts5
news.ycombinator.com 2y ago

Ask HN: Resources on Modern Cryptanalysis?

Al0neStar
1pts0
www.nature.com 2y ago

Three reasons why AI doesn’t model human language

Al0neStar
2pts1
osf.io 2y ago

Why agency and cognition are fundamentally not computational

Al0neStar
1pts0
marginalrevolution.com 2y ago

What (book) should Stripe Press reprint?

Al0neStar
1pts0
bnnbreaking.com 2y ago

Nvidia CEO Foresees AI Competing with Human Intelligence in Five Years

Al0neStar
3pts2
www.youtube.com 2y ago

Introducing Visual Copilot – Design to Code in a Click

Al0neStar
2pts0
akjournals.com 2y ago

Tescreal hallucinations: Psychedelic and AI hype as inequality engines

Al0neStar
2pts0
proton.me 2y ago

The Online Safety Bill looks set to shatter privacy in the UK

Al0neStar
7pts1
browserbook.substack.com 2y ago

Web Browser Engineering Concluded

Al0neStar
3pts0
www.youtube.com 2y ago

Graphically Impressive PS2 Games [video]

Al0neStar
2pts0
arxiv.org 3y ago

On the Paradox of Learning to Reason from Data

Al0neStar
2pts0
sites.socsci.uci.edu 3y ago

Poverty of the Stimulus Without Tears [pdf]

Al0neStar
2pts0
letterboxd.com 3y ago

Good Films, 90 Minutes or Less

Al0neStar
1pts0
gist.github.com 3y ago

Large Language Models and the Information Ecosystem [pdf]

Al0neStar
2pts0
www.science.org 3y ago

Art and the Science of Generative AI

Al0neStar
2pts0
www.tedsanders.com 3y ago

Predictions about the Year 2050

Al0neStar
2pts0
www.reuters.com 3y ago

U.S. Congress to consider two new bills on artificial intelligence

Al0neStar
2pts0
davidgerard.co.uk 3y ago

Crypto collapse? Get in loser, we’re pivoting to AI

Al0neStar
63pts53
www.springer.com 3y ago

Springer Book Series: Problem Books in Mathematics

Al0neStar
3pts0
www.artisana.ai 3y ago

China's Video Game AI Art Crisis: 40x Productivity Spike, 70% Job Loss

Al0neStar
39pts79
arxiv.org 3y ago

Data Contamination: From Memorization to Exploitation

Al0neStar
1pts0
www.youtube.com 5y ago

South Korean professor’s self-driving car was decades ahead of the curve

Al0neStar
2pts0
en.wikipedia.org 5y ago

Primarily Obsessional OCD

Al0neStar
3pts1

These books are what i would consider the best books for self-study. (I have gone through many textbooks after i went back to school to study statistics.)

• Real Analysis: A Long-Form Mathematics Textbook

• Proofs: A Long-Form Mathematics Textbook

• Mathematical Logic Through Python

• A Course in Calculus and Real Analysis

• A Course in Multivariable Calculus and Analysis

• Differential Equations With Applications and Historical Notes

• Probability and Random Processes (Grimmet)

• Probability Through Problems

• Fifty Challenging Problems in Probability with Solutions

• The Simple and Infinite Joy of Mathematical Statistics

• An Introduction to Error Analysis

everything is compute.

Isnt this the problem OP is referring to? If everything is compute then no useful distinction can be made.

[NSFW]

"… I show them my prick, then what do you suppose I do? I squirt the fuck in their face… That's my passion my child, I have no other… and you're about to behold it."

- The 120 Days of Sodom, 1785

"My dick will go through the middle of boys and the middle of girls, but with bearded men it will aim only for the top."

- An epigram of Priapus from 1st Century AD.

Assuming that HumanEval is a good benchmark (it's not) and assuming you can naively scale under fat-tails (you can't) then according to OpenAI's own gpt4 report (someone did the math on hn/reddit where they reproduced the curve but i can't find the link) at 100x the training cost of gpt4 you will still have a 15% error rate on medium difficulty tasks.

Fixed it, i meant that both of the books mentioned are written by Harry Percival (and the 2nd was written as a sequel) .

I'm guessing this was submitted after the "Ask HN" about leveling up to a production python programmer and i'm surprised no one mentioned these books:

1. Test-Driven Development with Python

2. Architecture Patterns with Python

The 2nd one is the closest you're gonna get to a production-grade tutorial book.

Related to this topic, these resources by @dbeazley:

Barely an Interface

https://github.com/dabeaz/blog/blob/main/2021/barely-interfa...

Now You Have Three Problems

https://github.com/dabeaz/blog/blob/main/2023/three-problems...

A Different Refactoring

https://github.com/dabeaz/blog/blob/main/2023/different-refa...

His youtube channel:

https://youtube.com/@dabeazllc