N1H1L

37 ms·

331.
by N1H1L 6y ago
You are forgetting the Mongols. They said the Euphrates ran black for a month due to the books dumped into the river. The Siege of Baghdad killed all of what you described
332.
by N1H1L 6y ago
So from reading the article, what McKinsey needs to now do is to hire McKinsey to make the difficult decisions, and "rightsize" the company.
333.
by N1H1L 6y ago
What is Saudi Arabia doing in Yemen then? Or what has Pakistan done in Afghanistan? Why single out Iran?
334.
by N1H1L 6y ago
Even more important to understand that we humans as a species are extraordinarily poor at understanding complex systems. Larger, older organizations will become complex over time, and complexity increases at higher than linear rates. This i
335.
by N1H1L 6y ago
Incentives in my opinion leads to Goodhart's law [1]. In fact, I will go a step further - most incentives are chosen by well meaning people. There is neither malice nor incompetence going on. What happens is that every incentive can
336.
by N1H1L 6y ago
The recent opinion section, no less by the Editorial Board, at WSJ claimed that focusing on renewable energy sources led to the Texas power cuts [1]. While the paper, on the same day itself, reported the opposite [2]. Links: 1) https:&#x2F
337.
by N1H1L 6y ago
Is Yi Cui with an h-index close to 200 the best scientist then? Or Pulickel Ajayan? I worked with samples grown from both of them, so my opinion on them is mixed. They are drive-by shooters; they identify a hot area, publish some crappy but
338.
by N1H1L 6y ago
National Lab scientist here. Too much of a revolving door has led to most National Labs becoming like universities now. LBNL is now effectively a UC Berkeley division (check it, their Lab directors, associate directors, division directors a
339.
by N1H1L 6y ago
Randomization. This is the only strategy, in my opinion, that makes sense. Every year, a certain fraction of the grant money from NSF/NIH/DOE will be awarded randomly, based on a lottery. Your name is automatically entered into th
340.
by N1H1L 6y ago
Unless you are in a user facility, national labs are even more brutal than academia. Because your entire salary is soft-money funded, and AFAIK you have to bring in several multiples (~4) of your salary in grants to maintain your position.
341.
by N1H1L 6y ago
That's a LOT. US as of 2019 had around 70k charging points [1]. Which means globally the number is not above 300k. Shell, if successful will install more charging points than there are there today globally. [1] https://evado
342.
by N1H1L 6y ago
The fuel for the PHEV still needs to be extracted, processed and burnt. It's not non-existent. Secondly, the weight and complexity of a PHEV is comparable if not more than a pure BEV. So when it comes to lugging around stuff, PHEVs are
343.
by N1H1L 6y ago
This is exactly what Michael D Higgins, the Irish president accused European historians of. He calls it a feigned amnesia over colonialism [1]. [1] https://www.theguardian.com/world/2021/feb/11/irish-pres
344.
by N1H1L 6y ago
> The key point is that a BEV and PHEV can provide similar environmental benefits. Yeah, this is dangerously bordering on falsehoods.
345.
The Battery Is Ready to Power the World(wsj.com)
4 points by N1H1L 6y ago|1 comments
346.
by N1H1L 6y ago
Tell me about it. I had cudatoolkit installed, and then followed every single PyTorch installation instructions, and then still it insisted that I should get the CPU version, not the GPU one. I ultimately fixed this with going to the conda
347.
by N1H1L 6y ago
Technically Cython is also a answer
348.
by N1H1L 6y ago
The Ju of Jupyter is for Julia.
349.
by N1H1L 6y ago
dask in my opinion comes decently close to better parallel computation in Python, and being beginner friendly. It is immensely compatible with numpy and pandas , and can abstract away a lot of the parallel computation complexities whic
350.
by N1H1L 6y ago
Libraries in Python are hell. And Stackexchange can be almost pedantically useless in those things. For example let's take OpenCV bindings for Python. 3 separate libraries provide it, which all are called in as import cv2 Yes
351.
by N1H1L 6y ago
Julia in many ways was designed as the next Python. Their user base is growing steadily, and they may displace Python in many applications a decade from now.
352.
by N1H1L 6y ago
India has a far lower per capita incidence than most Western countries, even a place like Germany for example. And anybody who has been to India won't classify its culture as rule-following.
353.
by N1H1L 6y ago
This is fundamentally because the risk-reward incentive structure is absolutely broken right now. If your lab expects productivity and transformative science (basically if you get a Nature paper every month that will be good!) then someth
354.
by N1H1L 6y ago
I agree with Robinhood's fears on principle. What however riles up people (legitimately IMO) is that larger funds making equally suspect trades are basically let off with a slap on the wrist. A strong argument could be made that Tesla&
355.
by N1H1L 6y ago
Also how do you quantify a bad trade? How much lost money? I trade on Robinhood too (not GME though), and just two weeks back I lost money on Nio stock. Is that a bad trade? Should Robinhood have stopped me from buying that stock?
356.
by N1H1L 6y ago
By the way, Dominion sued Powell for a similar amount two weeks back too. And FWIW, there are absolutely ZERO Fox News articles on either lawsuits, while every other news org covered it. I wonder why
357.
by N1H1L 6y ago
Thanks for that name. I went down a rabbit hole, and found out that the double agent who blew his handler's cover - George Blake died less than a month back!
358.
by N1H1L 6y ago
So if foreign talent recruitment is espionage bordering on theft then choking up the immigration pathway for trained scientists to become residents is arson too.
359.
by N1H1L 6y ago
> The Thousand Talents program essentially a foreign espionage operation citations please
360.
by N1H1L 6y ago
Yes. And GlobalFoundries has basically bowed out of leading edge nodes, saying they won't have the R&D or money to go below 12nm

More ›