In physics, you typically learn about error propagation quite early in your studies.
If you make some assumptions about your error (a popular one is to assume Gaussian distribution) then you can calculate the error of the result quite elegantly.
It's a nice excercise to write some custom C++ types that have (value, error) and automatically propagate them as you perform mathematical operations on them.
Unfortunately, in the real world only very few measurements have a gaussian error distribution, and the problem are systematic (non-random) errors, and reasoning about them is a lot harder.
So this means that automatically handling error propagation is in most cases pointless, since you need to manually analyze the situation anyway.
It's funny that the original idea for HTTPS was that there should be private communication between clients and service providers, and it somehow got turned on its head and now its just private communication between you and Cloudflare, and they can see all the traffic.
We talk about end to end encryption all the time, but half the web is hosted by a single company with questionable ethics and everyone is like, we trust them! They write technical blog posts!
This guy tweets all day and the press talks about every little thing he ever does, there are hundreds of videos of him on youtube, dozens of his employees and ex-employees telling stories about him on the net. How much more do you think you are going to learn if you actually talk to him?
I would argue a lot of Musk haters are not getting a representative sample of his behavior, but rather only the most outrageous things he says and does, which tend to be the things that penetrate their bubble.
I don’t hate him. I think he’s a sad, small man who makes himself feel big by being a bully and faking technical experience and knowledge. I think his fans confuse being in the right place at the right time in possession significant capital as genius. I think he is an incredibly toxic public figure. I think farming outrage and feeding trolls is bad for everyone. I think Nazi salutes are bad. I think that lying is bad. I think abandoning children is bad. I think buddying up with fascists is bad.
So sure maybe I miss out on his generous acts but honestly he does enough bad that I don’t particularly care about any good he does. He’s only doing it for himself anyway.
When I was a wee laddie, my dad bought a Herb Alpert album. I liked the music, and decided I was going to play the trumpet. I took lessons and practiced for 5 years. Making a recording and playing it back, forced me to accept that I had no talent whatsoever for trumpet playing. But I did learn how great a trumpet play Herb Alpert was. His control and tone and musicality is stunning. I could never hope to play like that, but I could enjoy his music and bought all of his albums.
Sometimes I'll play one of his songs to a friend, and gush a bit on how good Alpert played. My friends would invariably look at me in puzzlement. They simply didn't know enough about the trumpet to see the skill and virtuoso genius in his playing.
I've never heard another trumpet player play the instrument that well.
You keep saying I’m in a bubble. I read the Vance bio, and the Isaacson one. I came away with very different opinions of him. At least the Isaacson book didn’t leave me with the taste of boots in my mouth.
Is this the thing about California high-speed rail? I looked into that when it came up on HN the other day, and concluded that Musk had basically nothing to do with the failure of that project. https://news.ycombinator.com/item?id=43299460
So -- Do you have actual solid evidence to defend this "kneecaping of public transit projects" claim? Or shall I assume that the rest of your claims are also liable to be based on half-truths and internet rumors?
In any case, I'm not "defending" Musk in the sense of saying he's a good person. I'm just saying people should try to see him accurately. If you judge people negatively for giving you any information that challenges your worldview, I'd say you're basically admitting that your worldview isn't likely to be very accurate. You can persuade yourself of anything you like, if you're selective in the evidence you admit.
Is the “nazi salute”, still a thing?
It seems like accusations involving it are increasingly used as a tool to delegitimize, because cmon.. you can’t agree with him, but that wasn’t a nazi salute.
I think you should focus on things you can enforce. For example, in Austria companies are required to provide certain information in job ads (eg. salary range, weekly hours, type of contract) and it's trivial to enforce because you can see if the information is there or not. I'm not sure if it helps with ghost job ads, though.
So I don't know what medical education is like in other parts of the world, but in Austria it involves a lot of practice. Doctors spend a lot of time practicing medicine under supervision before they are allowed to practice on their own. Specialists work as "assistant doctors" for a few years before they can open their own specialist practice.
It's not a question of theory or practice; you obviously need both to learn advanced skills.
Let's put it this way, if the Russians get my ID and a picture of my face and know my kinks and my religious preferences what is the worst that can happen?
Now, if my government knows everything there is about me and one day decides to crack down on dissidents or hand them out to another foreign power as it was done in WW2 with the Jews in the Netherlands? Well, that is on another level.
We have be down this road before. It never ends well.
Maybe it's been too long and people forget. My grandparents lived through WW2, from what they told me, the capacity for humans to inflict pain and suffering on other humans knows no bounds.
Technically yes but Switzerland does subscribe to most EU legislation in order to join the internal market. They're a lot more "EU" than the UK is now.
They generally agree to pretty much everything yes because they're afraid of losing internal market access. It just takes a while longer because they have to approve each new thing individually.
I do use Switzerland as an exit for my vpn though yes.
This cope is officially dead now, not even Proton is believing Switzerland anymore. The pressure has gotten enough that they had to freeze all Swiss investment and start the process of moving key infrastructure to another country (I don't remember which, but it's the one Mullvad is at).
Truth is, Euros don't care about privacy. The endgame will probably be to host this stuff in the third world or something, like pirates do
Switzerland created a new set of surveillance laws in January, that far exceed anything inside the EU. Which means that EU laws are irrelevant, when talking about a company inside Switzerland - you should be talking about what they actually use!
I don't understand this. There is a red light. You will have to stop. Why do people still cross a solid line (which will get you a fine if police sees you) to overtake me and then stop directly in front of me? It makes no sense.
It's even worse when they overtake me and stop in front of me at an intersection with two lanes where they could just stay in the fucking lane they used to overtake me.
I swear some driver are doing it on purpose because they hate everyone else on the road.
In general I agree that messing with ecosystems sometimes has unpredictable consequences.
In the case of mosquitos, though, they cause so much suffering, that it would be stupid to not work on eradicating them because of possible negative consequences.
We have to be careful, of course (widespread use of insecticides is a problem), but targeted measures are really unlikely to cause more harm than mosquitos already do.
If you make some assumptions about your error (a popular one is to assume Gaussian distribution) then you can calculate the error of the result quite elegantly.
It's a nice excercise to write some custom C++ types that have (value, error) and automatically propagate them as you perform mathematical operations on them.
Unfortunately, in the real world only very few measurements have a gaussian error distribution, and the problem are systematic (non-random) errors, and reasoning about them is a lot harder.
So this means that automatically handling error propagation is in most cases pointless, since you need to manually analyze the situation anyway.
reply