本帖最后由 匿名 于 2022-2-27 01:05 编辑 .
感谢提供录像。我去生成了一个自动字幕,这页幻灯片对应的内容是这样的(从第 20 分钟左右开始):
Now, why do we need to be be careful, in particular, how can we engineer a more responsible digital future? Well, we need to be careful about inappropriate generalization. Right? I mean, you can write an algorithm for anything and everything. The algorithm doesn\'t know whether it\'s about elementary particles or animals or people, but that should make a big difference. Right.
And there have been quite a few people who have been developing big data algorithms for elementary particle physics. And these algorithms are extremely powerful in terms of how much data they can handle. And some of these people have also got interested in social applications and societal applications. And so they take their algorithms and adjust them somehow to apply it to our society. That is also, in a sense, to you.. check 1point3acres for more.
And the question is, how far can you go and should we do this? Like, should you apply algorithm that has been created for many particle systems to, say, chicken or pig? Certainly with the use of algorithms, you can optimize farms. Right.
But then if you apply the same kinds of algorithms or slightly adapted ones to society, wouldn\'t be ended up in an animal farm society. It\'s quite characteristic, however, that many countries are applying big data algorithms to identify terrorists and criminals. By the way, there are a lot of false positives here. I mean, basically, false alarms of more than 90% of alarms are wrong. So they\'re kind of concerns about the quality of these algorithms. In the meantime, big data algorithms are also used in wordified unemployment, and they\'re used to running entire societies. Like I mentioned, the social credit score in China. And then we could end up with applying this algorithm that were derived from many particle systems to everyone.
The question is, should we do that? And if you ask, how should we do it? 然后回到第 7 分钟左右是提到中国的幻灯片:
As you probably read, there\'s a social credit score which is being tested in China And actually being rolled out. China right here. People are being surveilled with a lot of cameras. I think China wanted to have about 500 million cameras and this is using face recognition and hundreds of other algorithms to basically monitor, judge, and control the population. So everything that people do or don\'t do might get plus or minus points. And the resulting social credit score would determine the kind of jobs one can get, the countries one can travel to and even whether you can still take the plane or a fast train and also the speed of the internet connection.. ----
So this is pretty invasive, one could say. And some people call it technological totalitarianism. 于是大概理解这老师为什么这么固执了,人家就是想黑一下中国政府,觉得黑中国是自己的自由。不过,我同意 LinkedIn 上一位同学的观点。你想黑政府也可以,但是为什么要开地图炮,把中国人都带上呢?如果支持把大数据应用在社会上就会变成 animal farm society,那么是不是用了 social credit score 的中国就是 animal farm society 了呢?虽然 Animal Farm (https://en.wikipedia.org/wiki/Animal_Farm) 是一本讽刺小说,不一定是说中国人是动物。但是平时不小心踩到别人也要说一声 sorry ,既然出现了误解,为什么不能澄清一下,还坚称误解没有关系呢?-baidu 1point3acres
而且,别的国家也有信用分,前面也提了一嘴 CIA 的事(第一分钟):
For example, CIA Director said, you\'re already a walking sensor platform. So computer, smartphones, iPads and so on, and contain a lot of sensors, and they\'re collecting data, by the way. Also, he said when your smartphone is turned off, it doesn\'t matter. The CIA can get the data. And he also said that the CIA was very nearly within the possibility to compute on all human generated information.. 1point 3acres
It\'s very hard to imagine how much information that is. But I recently read and not entirely sure that\'s true, that the NSA had 140 terabytes of storage space for every single person on Earth. 然而后面就只说中国了。这不就是在暗暗加强观众对中国人支持集权的偏见?
最后,建议大家也去看一看录像:https://video.ethz.ch/lectures/d ... d-248ca9b4e1a7.html
|