"The best minds that I know are thinking about how to make people click ads."
There is a considerable number of exceedingly intelligent people in pure mathematics and (theoretical) physics that don't work for Google, Facebook etc. and never would. Those that do often have nothing to do with Ads even three to four edges removed (think Martinis or the people at MSR).
As someone that has invented novel algorithms I still feel Patents are a net negative especially for software. I just wish EU would be a better place for „startup“ innovation. My experience with the university infrastructure in that area isn’t great:
Propping up China vis a vis the Soviet Union and Japan was a smart move. The State Department should have dropped them like a hot potato after the Japanese economy stalled in the 80s and the Soviet Union collapsed.
Go to a very expensive restaurant, your mind will be blown how good meat can taste. Food with lots of spices tends to hide the quality of the raw ingredients, but on their own tofu and meat are no competition in terms of taste.
I sometimes think that C++ is evaluated unfairly because it has a very fragmented ecosystem and painful package management. This leads to an enormous amount of code duplication and stylistic conventions, which further errode how pleasant it is to reuse code. It also makes it harder for the community to develop a common “taste”.
Playing against a lot of AI opponents is part of the fun. I would start on some easy mode on a standard map with the standard number of opponents and city states . That will be much more representative of the actual game experience.
I think that has reversed for Europeans at least somewhat. I’ve talked to a German very successful Havard Professor, who considered a move back to Germany because of the situation in the US. I personally would have loved to work in the US at some point in the past, now I am not sure it is worth it.
That is because there are no weights to begin with. In actual biological neurons synapses perform a rather complicated biochemical dance involving vesicles, ion channels, second messengers, whenever an action potential arrives. Some of that can be described by simple ODEs which might contain a constant that you could call “weight” but the actual physical synapses know nothing about that.
Store of value == be able to bury it in a grave for 5000 years and be sure that it will have kept its value when someone finds it. I am almost certain that bitcoin wouldn’t survive societal collapse.
Qatar is a US protectorate, just as it was a British before (https://en.m.wikipedia.org/wiki/Al_Udeid_Air_Base). The British still maintain planes there. They have no independent means of survival. Most of the oil and gas is either in the US / British sphere of influence. The few exceptions (Russia, Azerbaijan, Iran, Lybia, Syria, Venezuela), Transit countries like Georgia, Ukraine. All were messed with in subtle and not so subtle ways
A German, French and Russian alliance is probably the winning move and was thwarted more than once by Anglo-Saxon intervention. Right now the US actively tries to sabotage the Nordstream 2 completion (https://www.dw.com/de/ist-nord-stream-2-noch-zu-retten/a-538...), which would make Germany energy independent from US controlled Gas.
I think GP meant to imply that parts of the EU are effectively still US client states and not totally void of US influence on their politics. It's pretty common for top politicians in Germany to be part of some transatlantic organisation set up by the US ("Atlantikbrücke", ...). Same goes for journalists in leading news papers, which will result in more favourable US press coverage.
Well Tesla's sentry mode and sensor suite that is connected to the internet could actually pose a massive security risk. I wouldn't be surprised if they were banned from entering military installations and other sensitive areas, not that Germany has that many of them that aren't open to US military personnel.
It is relatively trivial to hook any new accelerator you develop into the popular deep learning frameworks. In the case of AMD there actually already exists a mature compiler framework for their GPUs and their cards are mostly on par with Nvidia's. Most deep-learning researchers don't write custom CUDA kernels, but simply stitch high level operations together in python. So as soon AMD delivers a performance / power advantage there will be almost no friction to deploy a AMD only cluster.
One of Nvidia's actual moats is their system building competency, which AMD lacks. They can sell you a box / a whole server room configuration, since their acquisition of Mellanox together with network equipment.
The stock price is in part determined by the discounted future Cashflow, the one of AMD is pointing upwards that of Intel seems neutral or declining. Intel had a long string of failures and bad acquisitions recently, which points to bad management. AMD has all the potential to attack Nvidia’s deep learning moat with in house talent. Maybe this is all wishful thinking though, but my AMD calls jumped 46% within a day.
Well that is what governments are for, no-one really wants to pay for tanks and aircraft carriers either. But politicians have managed to convince the public that it is in their best interest to do so.
Once they figure out UV, there are lots of in house innovations in the pipeline that could prolong Moore's law for the next 10-15 years. This is according to Jim Keller, who gave talks about how there could be up to 50x density increase. Most of that technology is proprietary and together with TSMC they are heading toward a quasi monopoly.
A good example of that is adding boost to your code base. Once you've added boost to implement one simple feature, eventually someone will use boost to do preprocessor magic or other meta-programming. Then there is no going back, because there is (almost) no way you will be able to reverse that decision.
At least in experimental physics a significant chunk of your time is taken up by menial duties. This might be babysitting the gas-detector or anything from server maintenance, PCB-design, ASIC development. Some of it can turn into your PhD thesis, a lot of it is just uninspired grunt-work you are expected to do. Of course you are also typically only paid ~65% of a full salary, so a significant chunk of your work is unpaid, because of course people expect you to work full-time.
An alternative way of stating this is that you are expected to do your PhD research in the 35% of the unpaid time.
This is in principle illegal, but because the institutes don't have any automatic time-tracking, you are instead expected to lie on your time-sheets and to underreport the time you've worked.
There is no recent history of the US interning and torturing parts of their own population because of their religion or ethnicity. That is in a totally different category than foreign policy or stuff they did while Mao’s policies starved 20 million+ people to death or student protests were suppressed by rolling tanks over them.
Well just in terms of chance of infection a mass gathering is much worse than individuals refusing to follow orders. We have those her e in Germany as well, including right and left wing conspiracy theorists.
And this is because protesting during a pandemic never leads to a rise in infections, so the thousands of BLM protesters surely had nothing to do with rising case numbers.
I agree and ironically I still find reading the paper edition of a high quality news paper to be much more pleasurable than scrolling through a news feed or navigating the ad-laden webpages of a newspaper. In fact the user experience on most of them is so much worse that it is almost laughable.
Also since the articles are laid out physically before you in a predetermined order, far more of your senses are engaged. You are also more likely to serendipitously read an article far from what you would normally be interested in.
As an outside observer most of American news reporting seems totally ridiculous. If the German government announced a retraining / job initiative there would be a factual article in serious print media. Any criticism included would most likely be a citation of an opposing party representative, labor or business interest. Opinions by the paper on the issue would be kept separate.
This article in contrast mixes opinion and negative rhetoric with facts. It reads like they want to make sure I will have a negative opinion of the program, while I typically don’t read the paper to be convinced of one thing or another.