qoez

11 ms·

151.
by qoez 1y ago
The counter argument is that the theoretical minimum is a few mcdonalds meals a day worth of energy even for the highest ranked human pure mathematician.
152.
by qoez 1y ago
This feels like a chatgpt summary which is against HN guidelines
153.
by qoez 1y ago
I'm going to allow this piece of news out of my control to ruin my day and leave me a nervous wreck for the rest of the week.
154.
by qoez 1y ago
I just use duckduckgo. Silly name but the results are no different (I switched not because of ai overview but because google said it had 'no results' for something that obviously did and I was fed up).
155.
by qoez 1y ago
Your mind hasn't adjusted yet to this common type of X vague-posting
156.
by qoez 1y ago
It's not really machine learning, it's just optimization (optimize the space of shapes to create the desired filter). The paper doesn't call it ML either
157.
by qoez 1y ago
I always had the impression it was the other way around, non autistic 'normal' people personifies non human objects. Anyway I always had a pet theory that the reason some people are fooled into thinking LLM text output is a real h
158.
by qoez 1y ago
Definitely responsible but transparent definitely not. Anyway he's probably just saying this because it's beneficial for nvidias bottom line if it's open.
159.
by qoez 1y ago
I was expecting it to ask you to say some bad words which also could work as a thing languange models can't do (unless u finetune of course but as a joke captcha).
160.
by qoez 1y ago
OpenAI uses scale via eg data made by humans through outlier. Part of this move could probably be seen as starving competition from data as much as buying talent at scale.
161.
by qoez 1y ago
The cynic in me feels like 'safer' is code for censorship in this case.
162.
by qoez 1y ago
Free game idea The Witness style game using this as inspiration for the map design. (The witness is already kinda like this but we need more like it, more games that's plesant just to walk around in even if there wasn't a game ele
163.
by qoez 1y ago
Safety research is mainly focused on avoiding bad words and introducing political bias. There's no fine tuning that can prevent mass unemployment and similar secondary effects.
164.
by qoez 1y ago
So far. Give it a few years when the core players have spent their way to market dominance and regulation kicks in and you'll see the price hikes investors have been promised behind closed doors.
165.
by qoez 1y ago
For me I did use them but it was during very early childhood so they have a special place in my heart.
166.
by qoez 1y ago
4 year old repo. (Though when I search it today I still see shirts with stripes.)
167.
by qoez 1y ago
I can't make up my mind if I'd prefer an AI boss or not. Human bosses can be quite terrible and not having to deal with an emotional being seems kinda nice.
168.
This year, I read ten important historical novels(twitter.com)
2 points by qoez 1y ago|0 comments
169.
by qoez 1y ago
Is it actually funny though?
170.
by qoez 1y ago
It's so easy to make bot accounts I wouldn't be surprised if a decent amount of them are manipulated.
171.
by qoez 1y ago
The Surveilled Child, The Surveilled Citizen, ...
172.
by qoez 1y ago
There's still way more signal there then really any other social network (even youtube is getting slop-y)
173.
by qoez 1y ago
I'm getting a phishing warning on that site but not on the one with properly spelled financial so I'm guessing the one you have is wrong
174.
by qoez 1y ago
There's no mention of llm or chatgpt or even ai in this https://news.ycombinator.com/newsguidelines.html
175.
by qoez 1y ago
That's a good argument, it's really making me reconsider previously held beliefs
176.
by qoez 1y ago
I only read the chatgpt summarized version of this but it makes decent steelman points against thiel (as someone who likes him and his ideas).
177.
by qoez 1y ago
The alt text does say "Also, I would be hard-pressed to find that wrench for $5." so I guess even at the time without inflation it wasn't really possible
178.
by qoez 1y ago
This is why AI safety is going to be impossible. This easily could have been a bad actor who would use this finding for nefarious acts. A person can just lie and there really isn't any safety finetuning that would let it separate the
179.
by qoez 1y ago
There must be some cool application for this but I can't think of what. I guess computing shadows and things like that but we often already have 3d buildings (though maybe not for rural areas like this does).
180.
by qoez 1y ago
Interesting reply from an openai insider: https://x.com/unixpickle/status/1925795730150527191

More ›