HN user

ftxbro

3,756 karma
Posts162
Comments1,971
View on HN
on.substack.com 2y ago

Substack Is Leaning into Politics

ftxbro
11pts14
twitter.com 3y ago

OpenAI Announces Goal to Solve Alignment of Superintelligence Within 4 Years

ftxbro
3pts10
en.wikipedia.org 3y ago

Muntzing

ftxbro
15pts1
en.wikipedia.org 3y ago

Bekenstein Bound

ftxbro
2pts0
www.youtube.com 3y ago

Lego Town Funk (AI video style transfer the music video is all Lego)

ftxbro
2pts0
noambrown.github.io 3y ago

Noam Brown of Libratus/Pluribus/Cicero Fame Has Moved from FacebookAI to OpenAI

ftxbro
2pts1
en.wikipedia.org 3y ago

Mathematical Universe Hypothesis

ftxbro
13pts3
twitter.com 3y ago

Paul Graham Observes That Wokeism Seems to Be on the Retreat

ftxbro
12pts18
twitter.com 3y ago

Paul Graham Quantifies Social Media Usage of 10th Grade Girls

ftxbro
2pts0
en.wikipedia.org 3y ago

Pigeonhole Principle

ftxbro
2pts0
en.wikipedia.org 3y ago

Normal Number

ftxbro
3pts0
en.wikipedia.org 3y ago

Ringwoodite

ftxbro
2pts0
www.theguardian.com 3y ago

Army of fake social media accounts defend UAE presidency of climate summit

ftxbro
4pts2
en.wikipedia.org 3y ago

Heavy Press Program

ftxbro
4pts0
en.wikipedia.org 3y ago

Rondel Dagger

ftxbro
16pts5
www.lesswrong.com 3y ago

Are we in an AI overhang? (2020)

ftxbro
1pts0
en.wikipedia.org 3y ago

Egregore

ftxbro
2pts0
news.ycombinator.com 3y ago

What Is a 'Dead Loop'?

ftxbro
1pts8
news.ycombinator.com 3y ago

Design a programming language for LLMs to learn?

ftxbro
3pts8
news.ycombinator.com 3y ago

Is Hacker News slow for you guys too? I think it's the Apple announcements

ftxbro
14pts2
arxiv.org 3y ago

Gorilla: Large Language Model Connected with APIs

ftxbro
5pts0
www.nationalacademies.org 3y ago

AI to Assist Mathematical Reasoning: A Workshop

ftxbro
1pts0
terrytao.wordpress.com 3y ago

AI to Assist Mathematical Reasoning: A Workshop

ftxbro
74pts4
en.wikipedia.org 3y ago

Opus Magnum (Zachtronics Alchemy Puzzle Game)

ftxbro
3pts0
en.wikipedia.org 3y ago

Zachtronics

ftxbro
29pts5
en.wikipedia.org 3y ago

Zaibatsu

ftxbro
4pts0
en.wikipedia.org 3y ago

Chaebol

ftxbro
1pts0
arxiv.org 3y ago

Cinematic Mindscapes: High-Quality Video Reconstruction from Brain Activity

ftxbro
1pts1
gwern.net 3y ago

A Personal View of Average-Case Complexity (1995) [pdf]

ftxbro
63pts3
news.ycombinator.com 3y ago

Poll: How many of you take as an axiom that a computer cannot be intelligent?

ftxbro
1pts4

It's not even clear whether two humans are experiencing the same flavor or taste when they agree that a food is "sweet"

OK but that's like criticizing every science result by saying "how did they control for the possibility that I'm a vat brain and nothing they did was real"

"AI" doesn't work reliably enough to be used for most stuff. Safety from that perspective is real.

    -: AI is unsafe because it isn't capable enough
    +: AI is unsafe because it's too capable

If someone wants the data but doesn't want to read a pdf wall, here are the important numbers:

The mothers of 54 non-autistic boys were surveyed, and 7.4% of them had at least one diet soda per day.

The mothers of 67 non-autistic girls were surveyed, and 17.9% of them had at least one diet soda per day.

The mothers of 140 autistic boys were surveyed, and 19.3% of them had at least one diet soda per day.

The mothers of 28 autistic girls were surveyed, and 10.7% of them had at least one diet soda per day.

They did some analysis that was supposed to control for income, education, ethnicity, and geographical location to get the odds ratios and confidence intervals that other people are commenting.

I wanna know: where's the AI advocates? The people who are speaking in favor of AI now are just doing so because they think it will make them rich. Where's the people welcoming the bot overlords? Why aren't they working to make our future be the shape they think it should be?

The call themselves e/acc. Chad effective accelerationists, making fun of the virgin effective altruism movement. They encompass a wide variety of degrees of irony like early flat earth society or early donald trump fans and they will probably end up similarly, where the movement becomes completely unironic and the ones who participated ironically get bored and do other things.

AI safety is some different things. One are the grifters trying to make money from fear. Others actually think that capable enough AI will irreversibly change our way of life in a bad way. For corporations AI safety just means brand risk and PR like not making another "Tay". For yet another group of people AI safety means the continuation of culture war stuff in the AI space.

I’m starting to think there’s an LLM equivalent to the old saying about how everything the media writes is accurate except on the topics you’re an expert in.

This is true for media articles but for LLMs I feel like it's the opposite. Like people who aren't specialists don't fully appreciate how great it is at those tasks.

My theory is that it's a kind of math that LLMs can do. Not that LLMs are necessarily doing it, but the same thing that makes it easy for LLMs also makes it easy for certain people. They can repeat words and phrases that they hear, and sometimes even rearrange them into new forms! One time some category theory guys were struggling with something and it was equivalent to topological sort in computer science. But the way they were failing at solving it sounded much more impressive!

I likely write a follow up post about my criteria and reserach and perhaps a follow up review if I end up pulling the trigger.

I feel like people who use the phrase 'pulling the trigger' in this way have an unhealthy relationship with shopping. They are shopping for dopamine and they are like the ones who have a hundred of the same hat and when they need more dopamine they buy that hat again.