Recent submissions

Comment · 2 months ago
"got hacked (in French)".

Got hacqued.

Comment · 3 months ago
From a post with "Tesla spyware" as title I would expect more than literally two sentences that say "[...] my Model X was uploading ~2.6gb of something to Tesla when it parked and connected to WiFi. Hopefully turning all these settings off stops it"
Comment · 3 months ago
Instagram's algorithm is based on what you interact the most with (even stopping to watch the reel, to read comments or to write one counts as interacting). You can start marking all that content as "Not interesting" and in the worst case just blocking those profiles to fix your feed and algorithm (or even go to your settings and tweak your interests directly).

But uninstalling the app is never a bad choice regardless

Comment · 3 months ago
For whoever is wondering why: bottom center 3x3 box
Comment · 4 months ago
Excellent choice, sir
Comment · 4 months ago
Comment · 4 months ago
Mine is currently Net, 7x7 grid with wrapping variant. I take about 5 minutes on average to solve a level, which is the sweet spot for me
Comment · 4 months ago
Maybe it's just a matter of how much effort people put into finding exploits on the two apps. If WhatsApp has many more (actual) users than Telegram, researching exploits on WhatsApp is more worth researching on WhatsApp than Telegram

A bit like how there's much more malware for Windows than there is for Linux

Comment · 4 months ago
> this isn’t a political fight. It’s not a he-said, she-said between tech billionaires. It’s a technical question.

> In transit. Between two online devices. With no cloud backup. With no business accounts. With no Meta AI features. With no linked devices. With no law enforcement warrant for metadata.

> Under every other condition — which is how most people actually use WhatsApp — the story changes dramatically.

Smells a lot like slop so I'll pass, no thanks.

Comment · 5 months ago
Thank you
Comment · 5 months ago
Breaking: language model whose purpose is to predict the most likely token, after being trained on non-uniform human-generated dataset, does not follow a uniform distribution.
Comment · 7 months ago
The higher the variety of notes (out of the overall 12 sounds in an octave) in the song, the less this becomes possible.

If your song is really simple, e.g. only consists of the 3 notes that make up a major triad (root, third, fifth), then this is definitely possible and you can just use natural thirds and natural fifths.

But as you start adding more notes, more chords and perhaps change of keys etc, it starts to break down.

That's the reason why J. S. Bach wrote The Well-Tempered Clavier. It's a collection of 24 preludes and fugues, in each possible major and minor key.

The basic idea was that if every prelude and fugue sounded good on an instrument (organ, harpsichord etc.), than it meant that the instrument was "well-tempered".

Using natural tuning instead of 12-TET would have resulted in some pieces sounding very good and other sounding very bad.

Comment · 8 months ago
I appreciate your good faith but I tried to copy-paste the first ~7k character of this (not yours!) article in an AI detector (gptzero) and it's "highly confident that this text was AI generated" with a probability of 100%
Comment · 11 months ago
Comment · 1 year ago
That makes sense, but I would expect this paradox to vanish (or at least get weaker) as you go higher in the hierarchy of technical positions (i.e. from junior to lead, to senior, to principal etc.). I would expect the workplace to somehow naturally get rid of the incompetent people, so that after a certain point you're only left with [qualified, competent] and [unqualified, competent]

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton