hiq

21 ms·

241.
by hiq 3y ago
> You can write a technically "factually correct" article by only including statements that heavily favor one side of the argument, which is what's done here. https://en.wikipedia.org/wiki/False_balanc
242.
by hiq 3y ago
Updated: https://archive.ph/WEoa2
243.
by hiq 4y ago
> You will see the 1,000 Indian engineers every year (let's say) moving to the US, but not the other 499,000 that doesn't or cannot. But why can't they? Isn't that mostly a visa issue that wouldn't apply to an Am
244.
by hiq 4y ago
I can definitely believe that some good European universities charge international students high fees that make them less appealing. My point though, is that not all of them do. E.g. the Computer Science department of ETHZ is usually ranked
245.
by hiq 4y ago
Why are rising fees not driving students towards cheaper universities rather than just no university? I understand that a US degree has more value than a foreign one in the US, but given the number of foreigners with foreign degrees settlin
246.
Lawmakers Trade Bank Stocks While Working on U.S. Bank-Failure Fallout(wsj.com)
2 points by hiq 4y ago|0 comments
247.
by hiq 4y ago
GPT doesn't do anything on its own, it needs a prompt. That prompt is basically code, written by a programmer, although it resembles natural language so we're closer to usable literate programming than we've ever been. Progra
248.
by hiq 4y ago
I guess the follow-up question is: if you do value these 100 units of work (1 employee == 1 unit pre-GPT), why did you only have 10 workers at the beginning? Doesn't it mean that 10 units of work is what you thought was optimal? What a
249.
by hiq 4y ago
Haven't most productivity improvements led to more rather than less programmers? Why would this one be any different? I was thinking today that with respect to programming, LLMs are a bit like spreadsheets. They make "programming&
250.
by hiq 4y ago
If a country has 100% of millionaires that inherited their wealth but haven't done anything since they're born, I wouldn't call it "productive".
251.
by hiq 4y ago
> Almost 10% of 400 million people are millionaires in the USA! Insanely productive all things considered. The number of millionaires is poor proxy metric for productivity, because of inheritance for instance.
252.
by hiq 4y ago
> something that was not clearly expressed I'd argue that in this case it's also in the interest of the community to downvote, since the comment adds more noise than it contributes to the discussion. In the end it's a rank
253.
by hiq 4y ago
I don't know about your comments, but if I see one which is easy to debunk by looking it up then I just downvote. There's no reason to be less lazy than the original poster just to prove them wrong.
254.
by hiq 4y ago
They have a bias towards big tech (more big tech employees submit their TC), so you end up with higher numbers. The question is whether you get a similar bias in every country. You compare the same company in different locations to avoid th
255.
by hiq 4y ago
> The median gets affected by it. A lot. You're saying it explains a multiplier of 4 between medians, do you have any source for that? If I look for e.g. Apple ICT3 in India vs the UK it's $58k vs $132k. https://www.
256.
by hiq 4y ago
https://www.levels.fyi/t/software-engineer/locations/india has a median in India of $26k. On the same website, UK and Canada medians are way higher, a little under $100k.
257.
by hiq 4y ago
Are you sure it's not because the prices on the Japanese website doesn't include taxes and surchages? Japanese website: > All amounts are in JPY. Taxes and surcharges are excluded. French website: > All amounts are in EUR. T
258.
by hiq 4y ago
> What is the issue with income inequality? https://en.wikipedia.org/wiki/Economic_inequality#Effects
259.
by hiq 4y ago
On the other hand I can usually come up with my own estimate of how trustworthy the answer is when a human gave it to me, e.g. thanks to: * their reputation with respect to the question domain (if I ask a basic C++ question to a C++ expert
260.
by hiq 4y ago
> 90% of the time people will use their laptops to look at instant messaging, forums, emails, or even try to get some work done. So your boss is not wrong. > even try to get some work done I assume it's not what you meant, but it
261.
by hiq 4y ago
It doesn't have to be true though, it just has to be a nice story. And this is the kind of interview questions you can prepare.
262.
by hiq 4y ago
> I've seen plenty of cheaters, they were the worst students In other words, you haven't seen the "cheaters" who were among the best students. "cheaters" in quotes because it's not clear to me that peop
263.
by hiq 4y ago
In Europe I still haven't seen a bill without the VAT clearly shown in the final bill. I've also seen shops showing the two prices, one without the VAT, one with. But if only one is shown, it should be the one I will pay.
264.
by hiq 4y ago
https://archive.is/20230211161656/https://www.ft.com/content...
265.
by hiq 4y ago
From the abstract: > This model's answers are preferred by humans 56% of the time to those of our human demonstrators, and 69% of the time to the highest-voted answer from Reddit. These are very interesting metrics.
266.
by hiq 4y ago
> Administrative violence? Reality distortion field? Extractivism? Note that the first term has good results if you look it up (one being "what is administrative violence"), while the two others even have Wikipedia pages.
267.
by hiq 4y ago
> I just update when I think about it. What about security vulnerabilities?
268.
by hiq 4y ago
I just pin to a specific code name (version) instead of a state (such as testing), e.g. right now I'm using bookworm which happens to be testing, and I'll still be using it for a while once it's been declared stable. If I fee
269.
by hiq 4y ago
I assume you're talking about: https://old.reddit.com/r/personalfinance/wiki/commontopics linking to: https://i.imgur.com/lSoUQr2.png
270.
by hiq 4y ago
> this is only because businesses have yet to be built around this technology No, these are fundamental problems that time can't solve. Businesses aren't built around this technology because it's not advantageous to use it

More ›