HackerLangs
TopNewTrendsCommentsPastAskShowJobs

ipdashc

732 karmajoined 5 tahun yang lalu

comments

ipdashc
·3 jam yang lalu·discuss
Given that fiber's been around for literal decades, though, and the Internet hasn't recently gotten more popular or anything, why would this suddenly have changed? I could believe what people are saying re. Starlink providing competition and finally incentivizing fiber buildouts
ipdashc
·4 jam yang lalu·discuss
Meh. It's one megacorp stealing stuff from another megacorp, hardly "appalling", who cares. I'd probably react the same way; I just wouldn't leak it to my next employer, that's dumb.
ipdashc
·15 jam yang lalu·discuss
A lot of people are defending this as similar to Old Linus, and while both (a) I kinda get it and (b) Linus stopped doing that because it was bad... At least when Linus did it, it was vaguely in the service of giving people a wake up call about their code quality so their future contributions are better.

Versus here, I'm not sure I see the point? I'm not up to date on this drama, but like, Bun has left the project/community. There's nothing lost from staying quiet or just going "okay, bye". Why is there a whole post about how shitty the guy is, what purpose does this serve other than dunking on him for leaving? It comes off as bitter and petty, not because it's wrong or the tone or whatever, but because there's no reason for it.
ipdashc
·3 hari yang lalu·discuss
Yes, I mentioned that. And especially for coal, it's a big issue. But realistically coal is dying anyways, it's natural gas that's the main competitor. And oil&gas:

- ... often has its land use sited far away, in Texas and the Dakotas, or even offshore. Solar would realistically have to go everywhere; even assuming perfect transmission(!), you don't want to bunch it up in one region so it's not a single point of failure.

- ... employs lots of people / requires constant high-paying labor. Solar is built once then mostly just sits there. Which is good! But explains why residents might be fine with the former, but not the latter.

- ... is more easily hidden away. See LA's hidden oil rigs.

- ... already kinda exists. Which is a lame argument, yeah, but obviously people are going to be fine with the pumpjack or oil field that's been there their entire lives, vs. new solar farms coming in and taking over what was forest or farmland.

Again, I'm on solar's side here, obviously it's a better land use. But we can't just handwave away the entire land use issue, it's that kind of stance that makes people grow hostile to it.
ipdashc
·4 hari yang lalu·discuss
I'm a big fan of solar but this seems like a stretch. Generating the same amount of power as one fossil fuel plant requires a vast amount of land area, so of course that's why people complain about it. Coal plants and mines can be pretty large but natural gas plants, the realistic alternative, are pretty compact. I'm not sure how the size of oil/gas fields compares to a similar amount of solar, but they are generally sited far away, where people don't see them - and can be built relatively visually unobtrusively (see LA). And then there's nuclear plants, pretty big, but putting out multiple gigawatts.

Again, I think fossil fuels need to go and solar is our best bet, but no sense in hand waving the entire argument. They need a lot of land area.
ipdashc
·8 hari yang lalu·discuss
In fairness, the phone network practically took a century to get fully set up, was quite expensive in its time, and it's pretty easy to run a dinky little wire pair to a house (sometimes they'd literally use barbed wire fences). And labor costs were a lot lower when the phone network was built out.

Electricity is genuinely hard and expensive, but we've accepted it as a basic need of modern life. Lighting, fridges, HVAC, kitchen appliances... People die if the electric grid goes down for long enough, but phones and Internet are a bit more of an optimization, a luxury.

Not disagreeing with your sentiment, though, just saying the scenarios are a bit different. Electricity and phone aren't "very easy", we just accept one as difficult but required, and the other (if you mean landline service) was already there and isn't really being maintained anymore.
ipdashc
·8 hari yang lalu·discuss
> I am clearly in the minority here but I have driven cars with and without CarPlay and honestly do not care whether a car has it.

I'm in the same boat, though with Android Auto or whatever. I'm honestly surprised how many people seem to need CarPlay; a comment or two down there's a stat claiming 79% of buyers wouldn't buy a car without it. Is it that different from the Android implementation? Is there something special about it?

I don't get it. It's a nice to have for sure, but honestly not that special, and gets annoying at times (when Android Auto connects on my phone, it tends to stop me from using the maps app on the phone, plus a few other minor grievances). And it's not much easier than just plugging in an aux cord.
ipdashc
·8 hari yang lalu·discuss
I feel like this argument doesn't work in all cases, because of course there are densely populated parts of America. Massachusetts is 7,800 sq miles of land and 7 million people. New Jersey is 7,354 sq miles and over 9 million people, so almost double the density of Switzerland!

So sure, nationwide policies in America have to account for all the empty space, but there's also wide swaths of the country that have relatively normal (if still overall low) levels of density. What's stopping MA or NJ from starting a similar scheme to the one in the article? Probably a lack of funding, state capacity, and political will.

If anything, the comparison probably falls apart because Switzerland is filthy rich. Apparently their GDPPC is $126k vs America's $94k, and crucially, I suspect the former is much more evenly distributed. All I have to go off of is visiting once, but it's certainly a very expensive and well-maintained country.
ipdashc
·9 hari yang lalu·discuss
Group chats have kind of taken on this role, haven't they?

As you mentioned, Discord fills that spot for young people, but in general I get the impression people spend most of their time on group chat/private server environments nowadays. Social media is mostly treated as read only, a place you get memes or news from. Maybe there's that one rare friend who actually posts on Twitter or Reddit.

This gets mentioned occasionally, but I'm kind of surprised how little people talk about it, still. All anecdotal for me, of course, but still I find it interesting.
ipdashc
·9 hari yang lalu·discuss
To add, a lot of the most infamous Internet subcultures, with the worst effects on the "real world" (think incels, /pol/ types, mass shooters) have their own forums or hang out on places like 4chan/8chan. Where there's no big tech corporation or finely tuned engagement algorithm. Just broken people finding a place where they can reinforce each other.
ipdashc
·9 hari yang lalu·discuss
Sure, nobody's saying Facebook is the good guys, but this thread is about addiction. My point is that a lot of Internet addiction happens just fine on its own. As bad as Facebook is, it would still happen without them, including if they were regulated in whatever fashion.
ipdashc
·9 hari yang lalu·discuss
> ... because bullying engagement makes money for big commercial social media sites

Teenagers have been bullying each other for literal millennia. Kids got bullied on IRC and chatrooms; they get bullied today on Roblox, on SMS group chats. None of these have engagement algorithms.

Again, the corporations sure aren't helping, but we need to confront the fact that a large chunk of this is human nature. I'd like a better Internet landscape but switching from Facebook to Mastodon or whatever, while a good step, isn't going to cure all ills. A lot of people in these threads act like every Internet toxicity symptom ever arose because of evil corporations.
ipdashc
·9 hari yang lalu·discuss
Yeah, this doesn't get brought up often enough. Honestly, while I don't necessarily disagree with it, I think a lot of the "social media is evil because it was engineered by psychologists to be addictive!" narrative is complete cope.

Cope, specifically, because we don't want to accept the fact that this kind of stuff is addictive on its own, that we are our own worst enemy; bot armies, evil corpos, and engagement algorithms don't help but they're not required. (That is, between your two theories, I think both contribute but it's more so the latter.)

I'm a pretty easily distracted person. I don't use social media at all. Yet I've been "addicted" on some level to this site, to news sites, to browsing Wikipedia, to traditional forums. So have plenty of others.

People don't want to face the fact that humans just really like having a giant source of stuff to entertain ourselves with, and are easily drawn into online arguments. Getting rid of the corporations and such would probably make a better Internet, sure, but it's not going to cure everyone's addiction.
ipdashc
·16 hari yang lalu·discuss
> It seems that you simply took the "hyped headlines" for the whole of the work.

Well, yeah, that's what I'm saying. It's odd that there haven't been any major headlines (customer interest, competitors' announcements, etc) other than their initial demo. Good to hear it's being worked on though!
ipdashc
·16 hari yang lalu·discuss
> Good models will require multiple Taalas chips

I guess that makes sense. Is this feasible, or does the added latency between chips kill any of the performance gains?
ipdashc
·16 hari yang lalu·discuss
In the sense of interested customers, online discussion, other companies doing the same thing, etc. Of course it takes time to get actual results, but from an outsider's perspective it's surprising that it was basically just their initial demo and that's more or less it so far. Excited to see if they come out with something this summer though!
ipdashc
·16 hari yang lalu·discuss
> 17k tps taalas chip

It's odd to me that I haven't heard anything about this approach (baking LLMs/weights into silicon directly) since. It seems almost common-sense that we're going to end up there eventually. And it feels like that point is drawing ever closer now that model capabilities, if not quite plateauing out, are at least getting to a "good enough" point for a LOT of use cases.

I wonder if it's being worked on in secret, if there's something about it that makes it infeasible, or if companies are really too nervous to lock in one model like that because the next one down the line could be a huge improvement. Re. infeasability, I have heard that the Taalas demonstration chip ran Llama 3.1 8B (a pretty horrible model) and that even that took a massive amount of transistors / die area. So it might just be the case that the good models are too big to fit on silicon?
ipdashc
·22 hari yang lalu·discuss
That makes a lot of sense. I didn't think about it that way.
ipdashc
·22 hari yang lalu·discuss
That makes total sense actually!
ipdashc
·22 hari yang lalu·discuss
I'm aware, I don't kill spiders if I can avoid it. And I know cockroaches are nastier. I just think it's surprising that people are so visually afraid of them, since they're not a very scary-looking bug.