Recent submissions

Comment · 6 days ago
The main reason I don't see it in production code is you significantly increase your chance to write seg faults that kills the runtime / behaves unexpectedly to surrounding systems
Comment · 2 months ago
Yeah down in Australia. For a few hours since 4am UTC or so.

Only seems to be Fable and Opus. Sonnet is still working fine.

Comment · 3 months ago
I was being a bit cheeky, but I’m not really arguing that individual inventions can be determined as good or bad. My point is it comes from the same underlying mode of production. "Claude is useful" and "the way we have organised society that led to its creation may be ecologically catastrophic” can both be true.
Comment · 3 months ago
Darn, I tried this and the lithium battery kicked in.
Comment · 3 months ago
The benefits of modernity (electricity, cars, iPhone, Claude) are good, but they come bundled with potentially terminal ecological costs which is bad.
Comment · 4 months ago
Writing a custom solver seems kind of absurd to me. There are 10s(/100s?) of open source ones out there. HiGHS in particular seems to be challenging the paid solvers with its LP improvements of late. But now I want to write my own.
Comment · 5 months ago
I disagree with the analysis, although I liked the opening paragraph. Why separate the economics of interest rates from the economics of AI productivity?
Comment · 6 months ago
Always good to challenge the narrative - but I don't pay for RDS Postgres because of the WAL, replication, all the beauty of pg etc. I pay RDS because it's largely set and forget. I am gladly paying AWS to think about it for me. I think at a certain scale, this is a really good tradeoff. At the very beginning it could be overkill, and at the top end obviously its unsuitable - but for most of us those tradeoffs are why it's successful.
Comment · 6 months ago
(disclosure: I wrote this)
Comment · 7 months ago
more throughput WITHOUT huge tail latency is my understanding. A user above posted this link https://lwn.net/Articles/994322/ which goes into the background. My mental model is "give the kernel more explicit information" and it will be able to make better decisions
Comment · 8 months ago
I've speculated about this myself, but haven't heard anyone actually discuss it or reveal/leak this is the case. Do you have a source for this?
Comment · 8 months ago
`peon --pack peasant`

This was my first thought too, thankfully they thought of it!

Comment · 8 months ago
This was a very bitter pill to swallow! It took me more than one mistake to learn this - "you are not the user".
Comment · 9 months ago
Not OP, but I think that without some creative impetus like 'agency', how useful is an AGI going to be?

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton