Hacker Newsnew | past | comments | ask | show | jobs | submit | zdragnar's commentslogin

Fantastic support can cover up a lot of sins. Brand loyalty isn't everything, but it is one of the reasons Apple stays relevant.

… or your preferred retailer. This story about Amazon just never gets old. Back in 2010—I believe it was—I ordered a 27-inch iMac. Over time, dust accumulated inside the display, so 18 months after buying, I filed a claim with Amazon. To my complete surprise, they sent me a shipping label, and just three days later, I had received a full refund of the purchase price.

The thinking goes new features typically means new customers, possibly new markets, as the product or platform expands.

To use the Netflix example: why does a company that ships DVDs need to add streaming to their website?


Nursing homes in the US have more variety of quality, but the issue of caregiver pay is the same.

Care paid for by the government is paid at very low wages, especially for those assisting adults with mental disorders who have behavioral issues. It can be a very physically and emotionally demanding job that pays peanuts since typically only a high school diploma or GED is needed as a qualification.

There's a lot of aspects of these jobs that a robot and AI will not handle well for many years.


I was obsessed with UT2K4 on PC and didn't have a console... The first time my friends got me to play Halo, I struggled so bad because it felt like everyone was moving in slow motion.

There was something just so satisfying with stomping someone with a manta, too, compared with the vehicles in Halo.

I don't remember if I ever tried 2k3, but I think I'd miss the vehicles on team matches at least.


> There was something just so satisfying with stomping someone with a manta, too, compared with the vehicles in Halo.

Felt pretty satisfying to do it with a Ghost as well...


It's worth noting that a lot of energy drinks have over 200, sometimes 300mg of caffeine per can. That means two cans can easily put you over what is considered to be the daily safe limit.

The source of the caffeine doesn't change the fact that it is still caffeine, and is still a stimulant. Add on top of that the very large quantities of B vitamins and other things that probably aren't great to overdo on a daily basis, and you've got a recipe for cardiovascular stress.


> It's worth noting that a lot of energy drinks have over 200, sometimes 300mg of caffeine per can.

This is somewhat untypical though; the most common variants are all around 320mg/liter or less (redbull, rockstar, most monster flavors, almost every soda).

Exceeding 400mg in a day when drinking nothing else is admittedly still pretty easy.


Every gas station and convenience store where I live has had half of their beverage coolers replaced with nothing but energy drinks. Red Bull, monster and Rockstar make up maybe a quarter at most of the shelf space used.

Brands like Celsius typically have over 200mg, where others like Aloha are 180mg. Starbucks triple shot coffee drinks are 225mg, and there's a fair few candy flavors like C4 Skittles that are all 200mg.

There's far fewer that are closer to the 300 range, I think Bucked Up and C4 ultimate get there but they're less common.

I've made a habit of looking at and remembering the caffeine content of drinks because while I can handle lots of coffee easily, the chemical cocktail in energy drinks means anything stronger than Red Bull is nearly guaranteed to induce anxiety.


Reign and I think also Bang have 300mg

> This is somewhat untypical though; the most common variants are all around 320mg/liter or less (redbull, rockstar, most monster flavors, almost every soda).

Those 5 hour energy drinks are almost 4000mg/Liter.

Basically, 230mg in 57ml.

I usually consume half at a time, but on long road trips I usually end up needing the 2nd half 2-3 hours later.


> Those 5 hour energy drinks are almost 4000mg/Liter.

Yeah but those are shots. Nobody has a casual glass of it.


Yes, but as the paper notes, concentration may be a factor, not just the absolute amount of mg.

White monsters are extremely popular and only have 150mg.

That’s actually lower than a plain black coffee at Starbucks


There's also the typical combination with taurine. It's possible that caffeine+taurine is more dangerous at the same dose than caffeine alone.

If the solar company doesn't, then it falls to the farmer to pay for it, and smaller than corporate farms will struggle to do so. The concern was even more valid back when panels had more toxic heavy metals in them, but it's still a valid concern.

> The CO2 vendors are really helped by it being invisible, aren't they?

At this scale CO2 is irrelevant. As an example, China alone burned more coal from 2020-2025 than the entire world did in the entire 19th century. None of the major emitting countries- especially the USA- have made any useful headway in reducing raw CO2 emissions, and sticking a small number of panels in rural Appalachia will do nothing to change that. It's wildly cost ineffective compared to the grid scale installs power companies are setting up of windmills and solar panels elsewhere.


A point I make is there is a lot of people pushing misplaced thrift. When you run the numbers the cost of a 100 acres of economically unimportant land is a small percentage of the cost of the solar panels to put on them. At the same time install costs are the lowest.

Big picture solar on houses, commercial buildings, parking lots is more expensive than just dropping solar on a 100 acres of flat dry scrubland.


Too bad people don’t seem to consider your argument and just downvote because you make some valid points. Dogmatism rules on topics like this.

How quickly we seem to have forgotten how often aws us east went down.

NL is, as you note, a tiny and very densely populated country.

For comparison, the nearest UPS distribution center from me is a 40 minute drive at highway speeds. Fedex is similar. The nearest USPS post office is only 10ish minutes at mixed speeds, but they have to receive something from one of their own distribution centers first, so same day isn't happening.

It's possible to get same day shipping from exactly one local store, and that's only for items they have in stock on the shelves that I could just go pick up myself. Next day shipping isn't available from anyone but a few big box stores, again for items they have on shelves, because the population density is low enough they don't send out trucks for just the one delivery.

OTOH, there's a lot of items on Amazon that are consistently two day deliveries (especially with Prime), because enough people are buying that they have at least one, if not more, deliveries from their distribution center to our post office every single day. It can get as high as 3 or 4 semi trucks dropping off packages at our post office during the holidays.

Edit: it's worth pointing out that I'm considered in an area that is only borderline rural, with my town having a density of roughly 90 people per square mile (34 per sq km). There's counties in my state with a population density of less than 1.


I read "more than one low end" as "more testing environments than one low end" rather than "more low end environments than one". It's a snarky way of saying they should have been testing high end environments for no more reason than they had the money to do it, rather than saying they should have been testing the low end more.

Is the speed due to a difference in the way it "reasons"? I was trying out a local version (3.8 27b) and watching the thinking, it's absurd how often it challenges itself. Simple facts get questioned multiple times, and it has to keep reminding itself that it already established some things as facts earlier.

Instructing it to ask any and all questions as necessary sends it into an existential crisis, because the question tool (oh-my-pi harness) suggested limiting the number of questions, and it couldn't figure out how to do both.


Aside from making our engineering prompt terse, I place a token budget on the thinking traces. Those traces are mostly meaningless(in the reasoning realm, not the usefulness realm), and just tell it to do something productive. For coding, this just makes it read doc, tests, etc.

I believe that's all reasoning levels do, but most inference engines should have thinking budget.


No it doesn’t look like it. In our data, Qwen Max uses about the same amount of output per task as Opus. So the problem does seem to be inference speed.

The issue you’re describing could be a small model problem


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: