Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
g-w1
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
1.
▲
by
g-w1
2y ago
I'm quite confused. In the article, the response from mathstral is also wrong???
2.
▲
by
g-w1
2y ago
> My deeper beef with this method is the complete absence of emphasizing, discovering or forming connections between cohesive things. We're trying to learn, it's a super power to start seeing patterns in what we learn, it forms
3.
▲
The AI 'Safety Movement' Is Dead
(bloomberg.com)
5 points
by
g-w1
2y ago
|
0 comments
4.
▲
Building intuition with spaced repetition systems
(jacobgw.com)
2 points
by
g-w1
2y ago
|
0 comments
5.
▲
What I learned from doing Quiz Bowl
(jacobgw.com)
1 points
by
g-w1
2y ago
|
0 comments
6.
▲
by
g-w1
2y ago
The Recurse Center[1] is the opposite of this. They don't take any of your income, they just help you find jobs and companies pay them to find cool people. And there is no curriculum. They believe that if people work on cool things the
7.
▲
Taking into account preferences of past selves
(jacobgw.com)
1 points
by
g-w1
2y ago
|
1 comments
8.
▲
$10M says we won't see human-superior AGI by the end of 2025
(garymarcus.substack.com)
14 points
by
g-w1
2y ago
|
13 comments
9.
▲
by
g-w1
3y ago
Yes, it's great. As long as you can follow logic and a bit of programming, you'll do fine. Just go slow (I had to re-read it a bunch).
10.
▲
by
g-w1
3y ago
You can use incremental reading, which is built on top of spaced repetition. Lots of people have invented it independently [1] and it works amazingly! Once you get the hang of it, it changes the way you think about learning stuff. [1] http
11.
▲
by
g-w1
3y ago
I'm not really sure what you mean? It's been empirically validated to predict how you remember better than alternatives (or equal to SM-17) (essentially when it predicts that your % recall is below a certain threshold, it just sho
12.
▲
by
g-w1
3y ago
Anki has literally changed the way I think. It's insane how I can just choose to remember anything and how I have gotten really good at creating flashcards to the point where I predict how I'm going to learn when making flashcards
13.
▲
A New Kind of Retreat
(recurse.com)
2 points
by
g-w1
3y ago
|
0 comments
14.
▲
by
g-w1
3y ago
With enough power, anything is renewable!
15.
▲
by
g-w1
3y ago
there are a bunch of pre-existing ones here: https://github.com/open-spaced-repetition/ > Multiple Language Support: fsrs.js, go-fsrs, rs-fsrs, py-fsrs, cljc-fsrs, swift-fsrs and ex_fsrs
16.
▲
by
g-w1
3y ago
Here is a question I asked about concrete examples of doing agentic things, because I think that concrete examples are necessary to expand your horizons: https://www.lesswrong.com/posts/umJmRfcJndY3Gsr36/concrete-e
17.
▲
by
g-w1
3y ago
Yep! I got some old word files off a floppy disk from 1996 and LibreOffice was the only thing I could find that could read them.
18.
▲
OpenAI Execs Invite Altman, Brockman to Headquarters in Push to Reinstate Them
(theinformation.com)
12 points
by
g-w1
3y ago
|
3 comments
19.
▲
by
g-w1
3y ago
There's a prediction market here about why he was fired: https://manifold.markets/sophiawisdom/why-was-sam-altman-fir...
20.
▲
The Tragedy of SBF
(scottaaronson.blog)
6 points
by
g-w1
3y ago
|
0 comments
21.
▲
The Puritans would one-box: evidential decision theory in the 17th century
(jacobgw.com)
3 points
by
g-w1
3y ago
|
0 comments
22.
▲
by
g-w1
3y ago
Yes, the only way to truly learn that you should never write a database is to write a database. In fact, I bet that is what the author of the tweet did to learn that they should never write one. I wrote a blog post about this: https:/
23.
▲
by
g-w1
3y ago
The shooting yourself in the foot part is doing memory unsafe things in C where you get a segmentation fault that Rust would have stopped you from doing.
24.
▲
by
g-w1
3y ago
g-w1.github.io/blog I love to write about anything that interests me. I want to become a better writer. I plan to become one through my blog.