Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
SmooL
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
61.
▲
by
SmooL
7y ago
You're also saving the time of network latency and waiting for a response. With the proxy, once the packets are out on the network, you're done and can move on.
62.
▲
by
SmooL
7y ago
I love the idea, and maybe this works well for them in a position of relatively lax performance. The truth is that all abstractions around data fetching are inherently leaky; making an API call does _not_ have the same runtime characteristi
63.
▲
by
SmooL
7y ago
Yeah my bad; my site is mostly just for my personal use and I didn't really optimize for views lol. I'm seeing about 200 views/minute.
64.
▲
My Second Phone Is in the Cloud
(lucassimpson.com)
192 points
by
SmooL
7y ago
|
103 comments
65.
▲
by
SmooL
7y ago
Maybe I'm missing something, but I don't understand the authors thesis here. He seems to have a problem with when he navigates somewhere and instead of receiving an expected block of pure text, he gets that along with a whole bunc
66.
▲
by
SmooL
7y ago
There's actually many games that account for bullet drop! The article talks about this, but off the top of my head there's PUBG, and any game that involves a bow and arrow as a weapon.
67.
▲
by
SmooL
7y ago
Java hate aside, I love the idea of pictorial format for specifying data flow. Haskell is wonderful in its partial application, '.' composing, and other such tools, that really let you quickly and easily specify how your data flow
68.
▲
by
SmooL
7y ago
Very impressive - key quote: "In addition, MuZero is designed to operate in the general reinforcement learning setting: single-agent domains with discounted intermediate rewards of arbitrary magnitude. In contrast, AlphaGo Zero and Alp
69.
▲
by
SmooL
7y ago
Amazing. Very cool to see this sort of emergent behavior. I also very much enjoyed this section: "We propose using a suite of domain-specific intelligence tests that target capabilities we believe agents may eventually acquire. Transfe
70.
▲
by
SmooL
7y ago
I always wondered if there was enough unsecured/vulnerable endpoints in the internet to provide enough compute "resources" for some sort of replicating script to survive. I guess botnets come closest.
71.
▲
by
SmooL
7y ago
"If you did those two things extremely well, the actual driving logic can be hardcoded." I mean, it _could_ be hardcoded, but there's millions of edge cases so it's pretty infeasible. I agree with parent comment that ful
72.
▲
by
SmooL
7y ago
Check out gold ETF's. They hold the physical gold so you don't have to, you just buy a stake in the physical assets. The ETF prices will vary depending on the specific ETF, but you can find ones that track very closely.
73.
▲
by
SmooL
7y ago
It seems since submitting that they are no longer the leader on the GLUE leaderboard - https://gluebenchmark.com/leaderboard/
74.
▲
by
SmooL
7y ago
I'm genuinely curious how you code in such a way that compensates for the problem. I'm unfamiliar with Smalltalk but very familiar with JS/Typescript. I'm sure that if tight-knit group held strongly to certain naming con
75.
▲
by
SmooL
7y ago
Well one thing: when you work on enough code for long enough you start to forget what's/what. Static typing let's you jump in and immediately know the shape of your data at any point in the code, without having to re-trace ex
76.
▲
by
SmooL
7y ago
As mentioned, the "25k/day" is for a future model in 2020. According to a linked video in the article, a human does about 10k/day. I believe that right now we're at the very start of this. Robots were supposed to co
77.
▲
by
SmooL
7y ago
Why does it matter where the energy comes from? Between natural gas, nuclear, and renewables, I see no reason to believe that energy costs will increase
78.
▲
by
SmooL
7y ago
Which has been shown time and time again to be the source of many serious software bugs and exploits
79.
▲
by
SmooL
7y ago
"-Normal distributions come from multifactorial situations, where many variables matter. -In a world governed by scarcity, our decisions and preferences tend to be multifactorial in nature. -In a world governed by abundance, on the oth
80.
▲
by
SmooL
7y ago
It's not full scale DotA (limited hero pool makes a huge difference imo) but it's still incredible. Obviously, this limited rule set is not pure DotA, and so the humans we're at a slight disadvantage. The results of letting a
81.
▲
by
SmooL
8y ago
As an aside, is no one else bothered by the ubiquitous null/undefined checks? Almost every function implementation here is started with checking that the args are actually what is expected. I get that typescript is only a compile time
82.
▲
by
SmooL
8y ago
I fail to see how they solve the central issue: validating that someone is actually a human and not a bot. Even the hash of the video of the father declaring the birth of his daughter; all it proves is that the video was _created_ before so
83.
▲
by
SmooL
8y ago
I think the authors point though is that all our effort into the algorithms where algorithms to do just one thing - search - and that we used that in conjunction with more compute power. I'll agree that the author emphasizes compute po
84.
▲
by
SmooL
8y ago
Or you're trying to do something fancy that the compiler/type system doesn't support
85.
▲
by
SmooL
8y ago
So it seems that the nest itself was secured with a repeated password, and there was no 2FA on their account (which they could, and were encouraged, to add). On one hand, it's easy to dismiss this as the nest owners being naive with in
86.
▲
by
SmooL
8y ago
Cartography is more than just land, it's about effectively communicating information that's inherently geographic. That can be cultural borders, political borders, shopping mall density, migration routes, ect. Infographics are tra
87.
▲
by
SmooL
8y ago
The right to self defense is indeed natural, nobody argues that. What I debate is that _method_ of such defense. Instead of weaponizing someone in a weaponized environment, does it not make far more sense to de-weaponize the environment? In
88.
▲
by
SmooL
8y ago
Not that I'm saying that what's happening is right, but rather just a serious question: why not move? I know it's not fair per se, but if it's within your means why not?
89.
▲
by
SmooL
8y ago
If you have so many papers to go through that you've started using such 'virtue signaling' as a means of filtering through them all, fine. But to apply such a method and then _take the time to comment about it_ completely def
90.
▲
by
SmooL
8y ago
Really? I get that it's not the norm, but to judge the whole piece _purely_ on the choice of units seems rather judgemental. The imperial system, while unconventional, still works fine.
More ›