Karrot_Kream

6 ms·

61.
by Karrot_Kream 4mo ago
See this kind of response is why online discussions of the topic become so silly. What is "broad nationwide"? What is "large portion"? How large should a "large portion" be? We have a history of blocking housin
62.
by Karrot_Kream 4mo ago
> The tariffs I understand (even if they really don't make sense in this particular case) but the permitting I do not. Do you have more information or links? Take a look at [1]. The current admin felt that NEPA reviews were taking t
63.
by Karrot_Kream 4mo ago
> I don't recall seeing Columbus as being significant on any US backbone map. More likely they just want to be close to each other. Someone must have started that ball rolling. AWS us-east-2 (2016) and GCP us-east5 (2022) are both i
64.
by Karrot_Kream 4mo ago
You're right. Of course political approval tends to involve a few more people than a couple dissenting commenters on a ratioed HN comment thread, depending on the regime in question. Pretty sure the entities in question had business li
65.
by Karrot_Kream 4mo ago
Elon has mentioned this in interviews and the S-2 for SpaceX calls out permitting challenges as a reason for moving to space. Elon being Elon, I don't really believe him but this is what he claims at least.
66.
by Karrot_Kream 4mo ago
And what happens when something you politically approve of needs power?
67.
by Karrot_Kream 4mo ago
The reason is because permitting and building a natgas generator is the easiest among the energy production methods in the US. Datacenters need to be close-ish to Internet Exchanges to be cost competitive when lighting up network capacity.
68.
by Karrot_Kream 4mo ago
In the past I wrote a version of HN that uses modern CSS rather than tables that populates stuff from the API. There I built a little blocklist of my own that prunes a comment tree the moment it encounters a blocked user (inspired by posts
69.
by Karrot_Kream 4mo ago
As an update to myself, the comments did eventually sort themselves out. I guess the initial "reaction" commenters and voters are just more interested in participating than in SNR. Good opportunity for me to finally start blocklis
70.
by Karrot_Kream 4mo ago
Okay (I disagree because most privacy policy discussions on HN go in the exact same direction and turn into outrage threads but this is a reasonable disagreement since not all of these discussions do), but model naming of all things? Come o
71.
by Karrot_Kream 4mo ago
If you have nothing valuable to say, don't say it? Not writing anything is a perfectly valid option.
72.
by Karrot_Kream 4mo ago
Seems like Fable is doing a lot better on SWE-Bench-Pro and FrontierCode than GPT-5.5. Given how most folks I talk to and people instead online keep mentioning that GPT-5.5 was better than Opus, I'm curious what the experience now is l
73.
by Karrot_Kream 4mo ago
I think the community on this site these days, much like other comment sections on the web, just read the headline and make a low effort comment. Regression to the mean I guess.
74.
by Karrot_Kream 4mo ago
That is 1000x better than griping about the privacy policy, capacity issues, token costs, and how trendy the names are for the new models (???). The bar is on the floor and I just want it at my knees.
75.
by Karrot_Kream 4mo ago
> Let me spin this a different way: Because there are no notifications, HN users need to frequently come back and check. This is "positive random reinforcement" and is one of the most powerful mechanisms known for creating and
76.
by Karrot_Kream 4mo ago
There were many attempts at this time to figure out how to scale forums. The bottleneck on forums and chatrooms was always human moderation. But a forum could only get so many mods and mods eventually burned out because moderation just real
77.
by Karrot_Kream 4mo ago
> I'm going to guess you're young enough that you don't remember when HN first appeared? FWIW this site has an older demographic and I doubt that's the reason. More likely most people weren't aware of HN until re
78.
by Karrot_Kream 4mo ago
My guess is this type of screen addicted user is overrepresented among internet commenters as only a small fraction of the population actually comments on anonymous/pseudonymous forums. Which is why it attracts so much energetic negati
79.
by Karrot_Kream 4mo ago
Andy Massey writes about this stuff and is generally heavily disliked by the anti-data-center folks * https://blog.andymasley.com/p/data-center-land-use-issues-ar... * https://blog.andymasley.com/p&#x2F
80.
by Karrot_Kream 4mo ago
The US has deindustrialized and since the rise of NIMBY politics hasn't built new industry in any scale at all. The first freeway revolts happened in 1959. In the meantime, basically everyone has agreed that industrial growth and manuf
81.
by Karrot_Kream 5mo ago
There was a minimum level of stability you needed in your life at that time to be able to own a computer and pay the phone bills needed back then (I came later in the mid 90s-00s but it wasn't too much different by then.) Kids needed p
82.
by Karrot_Kream 5mo ago
Maybe, maybe not. South Asia has a history of very weak government and I'm not sure it will ever, culturally, be able to stomach something stronger. From a Western perspective having experienced authoritarian states, it's understa
83.
by Karrot_Kream 5mo ago
If you want server side, as veqq says there's joy which implements its own lightweight http server. If you want a client, it's quite trivial to use the Janet C API to wrap libcurl or any other HTTP client. If you're open to L
84.
by Karrot_Kream 5mo ago
This is one of those arguments that is so vacuous you can apply it to anything and always be right. > "There's no way you'll hurt yourself walking to the living room" > "Read history: people always think every
85.
by Karrot_Kream 5mo ago
Have you actually put together trading strategies by having the agent drive? I've never trusted it that far and I use agents a lot at my job right now. The way I usually do it is, I break out pen and paper to do an analysis of what I w
86.
by Karrot_Kream 5mo ago
> my impression was that there are algorithms that work well (dont remember the name sorry), and they don't really keep track of the entire network per node, each node only has a heuristic idea of how to route packages both to local
87.
by Karrot_Kream 5mo ago
You pay interest on the margin you put up for shorts net profits from the position itself and cash or other assets you place inside investment accounts. You're also usually being charged interest at only a few basis points above the RF
88.
by Karrot_Kream 5mo ago
Heh if you look at the Sharpe Ratios of the strategies/models, they're all terrible. The best performing regime there is still a Sharpe of 0.019
89.
65% of Hacker News Posts Have Negative Sentiment and They Outperform(philippdubach.com)
6 points by Karrot_Kream 5mo ago|1 comments
90.
by Karrot_Kream 5mo ago
FWIW I built a text classification tool for internal use using (at this point 1 year old) frontier models and found that asking for reasoning significantly increased precision and recall.

More ›