3 ms·

No, the problem is that extremist content is a nonsense concept. Specifically, there's no reliable way of determining whether content is "extremist content" be

No, the problem is that extremist content is a nonsense concept. Specifically, there's no reliable way of determining whether content is "extremist content" because a lot of information that the political and religious extremes focus on is actually true information. It often these extreme views that provide unique insights and the most valuable information. So how - and moreover why - should Youtube try to sift through a physics lecture on metallurgy and attempt to differentiate it from a physics lecture on metallurgy. There's often no way of knowing whether the lecture is being given by a college physics professor or a 9/11 truther, except by examining the contents - and both might be telling the truth. Heck the college physics professor could be a 9/11 truther and the truther could be a college professor... that doesn't mean that this lecture on metallurgy should be removed from the college's account. The fact that some people can't tell the difference between an Einstein and a wannabe Einstein doesn't mean that we should censor them both.

WarOnPrivacy 6y ago
> There's often no way of knowing whether the lecture is being given by a college physics professor or a 9/11 truther, except by examining the contents Exactly this. Your points about extremist content stand as well. We've countless examples of NGOs that fight extremism, that get their content taken down due to algorithmic monitoring. AI aren't capable of this sort of nuanced decision making and there is no evidence that they're going to be capable in the near future. The only method to achieve consistent, quality content monitoring involves human eyes backed by qualified human judgment.
- goodluckchuck 6y ago
  That's the how, but not the why. A free flow of information is a net good. Restricting information is a net bad. Western society learned this 400 years ago. These are the same arguments which were used against the printing press, now being used against the internet, fighting against a new enlightenment the same way that the church fought to maintain their monopoly on knowledge. All combating "extremism" does is protect entrenched stakeholders from the truth. https://en.wikipedia.org/wiki/Sapere_aude https://en.wikipedia.org/wiki/Sapere_aude
  - WarOnPrivacy 6y ago
    A free flow of information is a net good - by default. From this point, publicly visible information that consistently shows to be counterproductive could be reasonably restricted from the highest volume platforms. 1 zillion loopholes ought to be enough for anyone.

WarOnPrivacy 6y ago

> There's often no way of knowing whether the lecture is being given by a college physics professor or a 9/11 truther, except by examining the contents Exactly this. Your points about extremist content stand as well. We've countless examples of NGOs that fight extremism, that get their content taken down due to algorithmic monitoring. AI aren't capable of this sort of nuanced decision making and there is no evidence that they're going to be capable in the near future. The only method to achieve consistent, quality content monitoring involves human eyes backed by qualified human judgment.

goodluckchuck 6y ago
That's the how, but not the why. A free flow of information is a net good. Restricting information is a net bad. Western society learned this 400 years ago. These are the same arguments which were used against the printing press, now being used against the internet, fighting against a new enlightenment the same way that the church fought to maintain their monopoly on knowledge. All combating "extremism" does is protect entrenched stakeholders from the truth. https://en.wikipedia.org/wiki/Sapere_aude https://en.wikipedia.org/wiki/Sapere_aude
- WarOnPrivacy 6y ago
  A free flow of information is a net good - by default. From this point, publicly visible information that consistently shows to be counterproductive could be reasonably restricted from the highest volume platforms. 1 zillion loopholes ought to be enough for anyone.