Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
acdha
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
34 ms
·
751.
▲
by
acdha
1y ago
Iran is a less sophisticated attacker and they attacked Israel on several days. Russia has been throwing everything they can make against the larger Ukrainian perimeter for multiple years. If Iran could keep pummeling Israel long enough for
752.
▲
by
acdha
1y ago
WebP produced good file size reductions because they recompressed JPEGs and ignored the loss in detail. I benchmarked it the day it launched and it was never once competitive enough to be worth the cost of using it. If browsers had supporte
753.
▲
by
acdha
1y ago
You’re talking about people who are breaking into U.S. citizens’ homes or vehicles and forcibly detaining them for extended periods of time or shooting them. There isn’t a magic phrase which helps someone who isn’t white and affluent get tr
754.
▲
by
acdha
1y ago
They asked for proof. You supplied evidence contradicting rather than proving the point. Now you’re trying to claim success by picking a statistical misrepresentation. Try to support the original claim if you want credit for meeting the ori
755.
▲
by
acdha
1y ago
> Counting the number of letters in a word is a significantly more difficult task than lying No, it’s not - you don’t even need to be literate to count symbols - but also consider the complexity of the second task and how many skills eac
756.
▲
by
acdha
1y ago
I think you could construct a stronger version of this complaint as FP languages not prioritizing usability enough. C was never seen as a great language but it was relatively simple to learn to the point that you could have running code doi
757.
▲
by
acdha
1y ago
That’s not inconsistent: your bluffer knows they’re making something up and is using their model of you to construct something they think you’ll believe. Someone who can do that isn’t going to suddenly forget how to count the number of lett
758.
▲
by
acdha
1y ago
The problem is consistency: AI tools usually produce output which _sounds_ like the top 10% but you have to read it carefully to find the bottom 10% parts. We’re not used to that because human performance isn’t that inconsistent and we use
759.
▲
by
acdha
1y ago
They claimed fraud by a single person, with zero replication. That’s both testable so they should be able to support it. At the very least, more than one researcher was involved and more than one ape was alleged to have learned ASL. There i
760.
▲
by
acdha
1y ago
So we’ve gone from “the U.S. carrying all of the water” to “the U.S. carrying 10 points more, in a way which heavily subsidized a key political group”? That seems like a pretty big shift in goalposts.
761.
▲
by
acdha
1y ago
We did get good returns but that was because other units did very well. It’s not a given but still too likely that they’d botch AI products but still be profitable due to ad sales continuing to grow. I do think AI has better than average ch
762.
▲
by
acdha
1y ago
> one piece of evidence I find very convincing is the observation that non-human primates exhibit the same sorts of gendered behavior with toys that human children do (females wanting to treat any kind of toy as a doll, males wanting to
763.
▲
by
acdha
1y ago
They didn’t lose interest in messaging when they cycled through half a dozen different apps but the problem was they were building products they wanted to exist rather than what users wanted. I’d worry about AI getting similar churn if PMs
764.
▲
by
acdha
1y ago
Also, I kinda feel dirty criticizing an open source project but Poetry seems to be struggling with technical debt. I hit bugs which have been open for years or stuff which is WONTFIXed, and while they truly do not owe me anything, it’s a lo
765.
▲
by
acdha
1y ago
You’re right in general but it has been striking to see how much the current Republican response has flipped from what we heard for years after the Elian Gonzalez case, where Giuliani described the BORTAC agents as “storm troopers” and the
766.
▲
by
acdha
1y ago
Faces can’t be easily be “faked” in person but they can definitely be misrecognized or recognized correctly against mistagged data. Anything like this needs to be designed with the assumption that errors will be frequent enough that the fin
767.
▲
by
acdha
1y ago
Yes, but ICE has been fast and loose with the law for a long time so they’re starting without something like the military’s culture about lawful orders, appropriate use of force, etc. As in so many areas, past presidents ignoring the proble
768.
▲
by
acdha
1y ago
Is it really a “back door” when it’s controlled by the network owner? It feels like we need a different term for that since it’s increasingly common on large networks.
769.
▲
by
acdha
1y ago
> If they are worried about being eventually targeted by the Trump administration, they have done more to paint a target on their back now then they ever could've done by quietly accepting the money. That's probably not true gi
770.
▲
by
acdha
1y ago
Exactly 50:50 doesn’t need to be the target but when you’re talking 80:20 on a large, lucrative field it’s more likely that there are interested people being deterred rather than just lack of interest. There certainly are many reports of so
771.
▲
by
acdha
1y ago
It’s really telling how they’re just so confident about easily debunked claims. I’m reminded of a retired college admissions administrator whose theory was that some of the men in college gap was over-confidence: statistically the women who
772.
▲
by
acdha
1y ago
I definitely agree it’s a problem. I haven’t seen any evidence supporting the idea that it’s innate rather than social in origin. Human neuroplasticity is our species’ big defining trait so it seems far more likely that we have a complex mi
773.
▲
by
acdha
1y ago
Speaking of politicized agendas, I note that you are asserting without evidence that they have a secret motive other than the one states while also assuming that the administration’s interpretation of the relevant contract language will be
774.
▲
by
acdha
1y ago
This is one of the great parts of the web: you can tag every element with the global lang attribute and have things work the way you expect. For example, you can have CSS generate the appropriate quotation marks even in nested contexts so y
775.
▲
by
acdha
1y ago
> Almost always when you take the top N of something, there is a higher proportion of males than the general set of participants. Try to support this with data and you run into some real messy analysis: in a few cases where testosterone
776.
▲
by
acdha
1y ago
They presumably didn’t write an entire literature survey in an HN comment because the CS pipeline problem has been written about so much in the past that it’s reasonable to assume basic familiarity: https://www.scientificamerican
777.
▲
by
acdha
1y ago
Where is your data showing those programs don’t exist? For example, conservatives like to talk about the plight of male nurses but even a cursory search shows that there are exactly the kind of programs you’d expect to find.
778.
▲
by
acdha
1y ago
No. I was just pointing out that your downplaying of the risks in this thread is too cavalier: I believe they think, as do I, that even the cost of testing the legality of a particular interpretation would be crushing for a small non-profit
779.
▲
by
acdha
1y ago
Exactly why they had to do this: the PSF mission statement is “to promote, protect, and advance the Python programming language, and to support and facilitate the growth of a diverse and international community of Python programmers.” Letti
780.
▲
by
acdha
1y ago
I like the bait bike operations some police departments do to catch the shops buying stolen bikes. Addicts steal things they can fence and cutting into the business side means you don’t have to catch nearly as many people, although Facebook
More ›