Again those findings: “We also find that, *on average*, the YouTube recommendation algorithm pulls users slightly to the right of the political spectrum”.
The whole “on average” nullifies the notion that occasionally recommending Young Turks to kids watching Anime once in a while somehow makes up for the fact that they push OANN or Newsmax even harder. That’s like saying I took one step forward so you should ignore the two steps I took backward.
Also you are ignoring the implications further down the line. If YouTube pulls neutral to the right, then it likely pushes those already right even further in that direction.
Are you familiar with the concept of network effect?
> So the indoctrination isn’t obvious? So it’s subtle? That makes it more pernicious, in my eyes.
>> No, that is not what the study found
“In my eyes” isn’t analogous to “that’s what the study found”, FYI.
I didn't say that they do it "as much", I specifically suggested they may do it half as often. But, per the study, they DO do it - otherwise, this would not have been a "slight" bias, it would have been a whopping huge bias.
What I meant to say was the assertion that occasionally recommending YoungTurks somehow mitigates the right-leaning bias of the site, as suggested with the statement “then it's not really a significant force in this area” is false. The site has a demonstrable rightwing bias.
Elections can and are decided by a few thousand or few hundred votes in battlegrounds. As such, the argument that it’s of negligible effect rings false to me.
Now, if I'm watching Anime and YT recommends Shapiro, I can agree that's closer to indoctrination. However, if it only happens like 2 times for every 10M watches of anime, and then 1 time for every 10M it's recommending Young Turks, then it's not really a significant force in this area; and it is only pushing slightly to the right - and I believe this is the sort of thing that the study found. So coming back to your first quote:
> So the indoctrination isn’t obvious? So it’s subtle? That makes it more pernicious, in my eyes.
No, that is not what the study found. It found that political recommendations for right-leaning content are slightly more common than those for left-leaning content.