Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jhrmnn
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
181.
▲
by
jhrmnn
6y ago
I wouldn’t say it’s confusing—it’s wrong and suggests at the least sloppiness in terminology
182.
▲
by
jhrmnn
6y ago
It could, but it isn’t in this case.
183.
▲
by
jhrmnn
6y ago
That’s not different from pre-Internet era, though. Subscriptions were never anonymous.
184.
▲
by
jhrmnn
6y ago
No, they are very crude (but useful!) models of reality. General relativity and quantum electrodynamics are much better corresponding models, respectively, and even those are just approximations.
185.
▲
by
jhrmnn
6y ago
By this metric, nothing has been ever solved in natural sciences. So this is not a useful metric.
186.
▲
by
jhrmnn
6y ago
Blogs are probably not a good example. There are so many useful services people are used to have for free on the internet, which will need to be paid without ads. Most of those could be subscription-based though. As for blogs—IMO articles t
187.
▲
by
jhrmnn
6y ago
Since when is constructive criticism expressed in words like “awful”?
188.
▲
by
jhrmnn
6y ago
Yes, conservation of mass is to conservation of energy what Newtonian mechanics is to general relativity. The former work pretty well in everyday life
189.
▲
by
jhrmnn
6y ago
You can be a capitalist and still say true statements about anarchism, both supportive and unsupportive. Or vice versa.
190.
▲
by
jhrmnn
6y ago
But Apple is more than a payment processor here, they are closer to a regular online store in this regard
191.
▲
by
jhrmnn
6y ago
I should have mentioned that I of course use Git. But the need to manually keep track of the calculation~commit pairing is tiresome and error-prone.
192.
▲
by
jhrmnn
6y ago
In all my papers the results were produced on multiple days (spanning months), with multiple versions of the code, and they are computationally too expensive to reproduce with the final version of the code. I'm trying to keep track of
193.
▲
by
jhrmnn
6y ago
Especially since there is no such thing as water memory.
194.
▲
by
jhrmnn
6y ago
Sound moves 4-times faster in water than in air, I assume even faster in ground surface, so let’s just assume it arrives instantly via ground. This would mean a delay like in lightning/thunder, which is what one can see in the video (2
195.
▲
by
jhrmnn
6y ago
But the heat capacity of the air inside is negligible compared to that of the solid stuff inside. So it shouldn’t matter if you do it just a few times in two weeks.
196.
▲
by
jhrmnn
6y ago
Have a Phd in physics, agree fully. I love talking to people about the physics and chemistry in every-day life, and you need very little of quantum mechanics or advanced electromagnetism. Perhaps statistical physics (thermodynamics) is the
197.
▲
by
jhrmnn
6y ago
Turning off a server on its own isn’t removing web history. Even google cannot be forced to host websites for free indefinitely.
198.
▲
by
jhrmnn
6y ago
The response of the immune system to the virus and its mode of operation (deadliness) may be entirely unrelated. The T cells may react to some parts of the virus protein, other parts may be related to its effect in the body once infected.
199.
▲
by
jhrmnn
6y ago
Interesting. Is this really not confusing? I haven't seen it done elsewhere. I find it confusing just because of that. Their reasoning for why they do it seems actually reasonable. EDIT: Actually the fact that their rule is followed by
200.
▲
by
jhrmnn
6y ago
I fully agree and try to do that in my research, but it's very hard to do 100%. I just wrapped up a 1.5-year computational research project, and we have all the raw and processed data, the main code, the processing scripts, the noteboo
201.
▲
by
jhrmnn
6y ago
I'm an author of one of the arXiv papers above. One thing to consider is that the approximative power of a given parametric function is not the only criterium. Being able to optimize that function efficiently is as important. Neural ne
202.
▲
by
jhrmnn
6y ago
IMO not a revolution, but I can see a solid evolution. My reading of the work on embedding ML into physical models so far is that the best strategy is to take it as far as possible with the standard physics approach of abstraction and reduc
203.
▲
by
jhrmnn
6y ago
I wonder if 10 years ago one would believe that we’d have a model capable of generating an article that fools many given a complex prompt, yet is essentially incapable of any reasoning.
204.
▲
by
jhrmnn
6y ago
It is, though. It doesn’t speed up function execution, but it literally speeds up the function call.
205.
▲
by
jhrmnn
6y ago
Very much agreed. I’ve once tried to count calories for about a month out of curiosity. Having Fitbit and tracking all my (amateur) cycling with a powermeter, I could record calorie output consistently. But if you want to eat out regularly,
206.
▲
by
jhrmnn
6y ago
Agreed. To put it simply, if you model a complex system with another complex system, of which you have a similar level of understanding, then even if the model is very faithful, you gain very little. Unless the model is transferable, so you
207.
▲
by
jhrmnn
6y ago
Agreed, it’s my go-to when I need a simple recipe for a common meal
208.
▲
by
jhrmnn
6y ago
It uses this under the hood https://www.macrumors.com/guide/exposure-notification/ So I believe it does work without internet connection
209.
▲
by
jhrmnn
6y ago
Unlike your comment?
210.
▲
by
jhrmnn
6y ago
What matters on curvy descents is how much speed you can bring out of the turn. And that mostly depends on how efficient lines you can pick.
More ›