Recent submissions

Comment · 3 days ago
I would say Google is buying shovels and selling shoveling services.
Comment · 1 month ago
How is Russia's army becoming stronger? They have lost so much people and equipment. Why would they attack EU when they can't defeat Ukraine after so many years? Are you suggesting they will perform a suicide attack against EU, cause as much damage as they can, and get absolutely destroyed in return?
Comment · 1 month ago
Let's say we were in an alternative reality were we had reached this quality of token prediction with just Markov chains. Would you argue that those would also be conscious? Or is the obfuscated behavior of transformers part of the possibility of consciousness?
Comment · 4 months ago
I am a dog owner, and I agree that a loose and aggressive dog is a liability to humanity and you are making everyone a favor if you kill it even if you could have escaped (e.g. you manage to get into your car in time). If it was attacking you, it could be a child (or another dog, too) next time. I would say that even a loose, non aggressive dog of a breed that is very statistically likely to mangle people (e.g. Bully XL) should be taken away from their owners that same day.
Comment · 10 months ago
My wife runs https://www.saviament.com/, an open-access educational website in Catalan. She also sells printable content following the same style as the website, which has exploded in popularity this year and has become a decent source of income.
Comment · 1 year ago
Wouldn't all the houses in the village lose way more than £400 in value?

Last I was comparing houses in a neighborhood, the houses near a powerline were consistently worth 15k-30k less (3-6% less).

Comment · 1 year ago
The post I was replying to was edited and originally claimed that London had higher murder rate than New York.
Comment · 1 year ago
You are right. Should not have relied on the most likely LLM-generated description attached to the data. I trusted it because I already had the wrong impression that it was less safe, but that was just because the raw number of crimes is high because it is very populated.
Comment · 1 year ago
London murder rate per million in 2024: 11.6 NYC murder rate per million in 2024: 43

And from what I read, NYC is exceptionally safe for a US city, and London is exceptionally unsafe for a UK city.

Comment · 1 year ago
Same boat. I have a 6D Mark II since 7 years ago now, and I misguidedly was counting on ML to be released within 3 years of my purchase. But luckily, it's still a fantastic camera.
Comment · 1 year ago
Can't the Western government strongarm the receiving bank by threatening to kick them out of SEPA/ACH/Fediwire?
Comment · 2 years ago
If you are going scam people with locked iPhones over the internet, why send an iPhone at all? You should just send a literal paperweight. It will serve the same purpose. It's not like the marketplace will be like "It's an iPhone. Even if it is locked, the seller is right. We will remove the negative rating and let the seller keep their reputation".

By the way, I recently sold an iPhone 8 (through App, but the transfer was in person) and out of 5 potential buyers, all of them asked the same three questions: "Is it on and unlocked?, is it carrier sim-locked?, which is the battery %?".

Comment · 2 years ago
It seems to me that it would increase safety, but I guess only a simulation could truly tell. I understand your take, but I have a different one: most accidents happens when someone makes a mistake at stop light changes. Less cars stopping/starting at red lights seems less chances for accidents to me.

Let's go with an absurd example: If you multiplied by 5 both the length of red and green stages, I would expect much less accidents per day, but it would obviously also be much more frustrating to move around (and possibly more accidents if the amount of infractions increase because of the frustration). If you divided them by 2, I would expect much more accidents of all kinds.

Comment · 2 years ago
I was trying to write a decent definition but Wiki does it better: https://en.wikipedia.org/wiki/Large_language_model#Agency

In this context it could be able to do multiple actions in order to address an ask from the user: read cells, documentation, edit cells, and perhaps even read the result from the edited cell before answering. i.e. "Can you create a new sheet that focuses on the Countries where the sales happened, including YoY differences, and tell me which countries are outliers and for which reason(s)?" (probably very far fetched given the level of progress Excel has achieved in 20 years).

Comment · 2 years ago
This is the source to the 3.9TW number: https://assets.bbhub.io/professional/sites/24/BNEF_2023-11-2...

I understand your frustration. I had to dig quite a bit to find numbers I could use, because there are multiple energy categories that are used inconsistently: solar&wind, all renewables, non-fossil. Then they are combined with two main ways of measuring them: capacity and average production. Then there are targets, predictions and in-production. It is a nightmare to get fair 1:1 comparisons.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton