Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
kofejnik
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
35 ms
·
361.
▲
by
kofejnik
8y ago
Like many things in life, it can be solved with money. So: 1. become a really good front-end developer, do remote work. Be very professional about it. You'll start making good $$$. 2. Search HN for tips on remote/contract work, th
362.
▲
by
kofejnik
8y ago
> If any of those folks who don't take the city up on the free apartment do something illegal, like defecating in public, they can probably be incarcerated or institutionalized for $30k per year. Should a non-homeless person be inst
363.
▲
by
kofejnik
8y ago
Exactly this. Literally anyone can randomly luck a few tourneys. Case in point - Main Event winners such as Jamie Gold and/or Jerry Yang.
364.
▲
by
kofejnik
8y ago
+1 for RQ, it's so much cleaner
365.
▲
by
kofejnik
8y ago
> such intimacy holds a special enough place in human psychology that's exactly what religious conservatives of all sorts say when they want to prosecute extramarital sex
366.
▲
by
kofejnik
8y ago
thanks for sharing, it seriously is amazing
367.
▲
by
kofejnik
8y ago
it's not survivor bias if you're studying long-term survival
368.
▲
by
kofejnik
8y ago
in 2007, it was a widespread analyst consensus that iPhone will fail, since competing phones have obv. better specs
369.
▲
by
kofejnik
8y ago
Nothing is guaranteed, however I can't see how this leads to US - China equivalency w.r.t human rights. > that's not the only thing that matters Maybe, but you prefer to ignore it altogether when equating China and US
370.
▲
by
kofejnik
8y ago
It's very easy to criticise US from the high moral ground of your home, where your safety and freedom are guaranteed by the same US of A. But let me ask you: would you rather be a dissident in China or an anti-establishment protester i
371.
▲
by
kofejnik
8y ago
tl;dr: CNN-based models learn to use scanner model and other external data to bias predictions (e.g., portable scanner correlates with higher likelihood of pneumonia)
372.
▲
by
kofejnik
8y ago
but the CTO is a nice guy who just had a management training, so instead, both you and Tim are sent to a mandatory conflict resolution class
373.
▲
by
kofejnik
8y ago
immediate_cognitive_cost(making a case for a raise) > immediate_cognitive_cost(doing nothing and hoping it sorts itself out and _maybe_ having to replace) but obv it changes when the departure looks imminent, as I said
374.
▲
by
kofejnik
8y ago
no, managers are not stupid, it's just that their incentives are aligned this way to get you a raise, (s)he needs to do a lot of convincing to their boss, HR, etc - a lot of trouble, really, so they are not naturally inclined. But havi
375.
▲
by
kofejnik
8y ago
I only eat from 4pm to midnight (wife hates it if we don't have dinner together), and it's not a problem to last till 4 on back coffee alone if you tend to sleep late
376.
▲
by
kofejnik
8y ago
You don't need to create any accounts with Samsung, they already have all your data anyway, linked to your device Id. Your data is safe in their cloud, don't worry. /s
377.
▲
by
kofejnik
8y ago
update: it has been fixed very quickly, and I'm a fan of triplebyte as a result
378.
▲
by
kofejnik
8y ago
update: it has been fixed very quickly, and I'm a fan of triplebyte as a result
379.
▲
by
kofejnik
8y ago
No, of course not. Once money exchanged hands, it's not their product, it's mine, as is _all_ of its output, including logs, etc. Otherwise, I'd call it 'spyware'.
380.
▲
by
kofejnik
8y ago
I believe GDPR does cover the security camera scenario, there was a comment here about a week ago from someone who successfully petitioned the German police to delete all footage of him at some protest
381.
▲
by
kofejnik
8y ago
No, it's my device that creates the data, so it is mine. Otherwise, does the typewriter manufacturer own the stories you type?
382.
▲
by
kofejnik
8y ago
For one, Facebook obviously had access to contacts and photos, without any way for user to disallow it.
383.
▲
by
kofejnik
8y ago
Well, I did email @support yesterday and also checked my spam folder, just in case. Still nothing.
384.
▲
by
kofejnik
8y ago
thanks, @support emailed
385.
▲
by
kofejnik
8y ago
>> a vector of your skills vs. a vector of job requirements Of course it's an over-simplification, but you don't think there's a legitimate relationship between the two?
386.
▲
by
kofejnik
8y ago
It works very nicely for predicting what movie you'd like to watch, so it could potentially work with jobs, too. Collecting enough data could be challenging, though
387.
▲
by
kofejnik
8y ago
> after the interview, everyone gets detailed personal feedback on every section, regardless of the outcome Wait, what? I never received any feedback _at all_
388.
▲
by
kofejnik
8y ago
Vector thing actually does make some sense, you might want to read about embedding. Or just watch fast.ai lesson 5. Ironically, this was the topic of my final (and failed) conversation with them
389.
▲
by
kofejnik
8y ago
Just my $0.02 about triplebyte - I went through all interviews, seemingly doing fine, but after the last informal talk (which also seemed ok to me), there was about 10 days of total silence. Finally, I emailed to enquire and immediately rec
390.
▲
by
kofejnik
8y ago
> I'd rather put two people in a run-down $1.5k apartment each than flip a coin, put one of them in a $3k apartment, and leave the other on the street. Do you want to get USSR? Because one committee plays God and regulates who live
More ›