Americans lose confidence they can sniff out social media bots
techxplore.com
techxplore.com
> why would a small percentage, though small, find anything positive about the bots, about being lied to
Under the "bots = automation" definition, this is similar to saying "why would anyone find anything positive about computer programs".
Instead, this article is saying:
bots = lies
Does the article actually talk about astroturfing trolls leveraging automation? The automation part doesn't seem to be relevant, as paid trolls are the problem.
In the end, I understand that the article has a problem with lying on the internet, but what does "bot" mean to the society at large these days?
Bots that attempt to make themselves indistinguishable from humans - the bots under consideration in this article - are attempting to deceive people. If something is signed 'Amazon Ad Bot' or 'Moscow Political Ad Agency Bot, funded by the Bob Charles for Senate campaign', that's one thing. If it's signed 'Amy Jones", that's a lie.
> The automation part doesn't seem to be relevant, as paid trolls are the problem.
Without the automation, the paid trolls' reach would be far less. It's like saying, 'the nuclear weapons part doesn't seem to be relevant; it's the hostile dictator that's the problem'.
So far, the most common issue raised in this HN discussion is the poor bots being demonized. Only on HN.
But as with hacker, these terms have basically lost all meaning now.
Hell, someone the other day passed me their phone and asked me to take a “selfie” of them...
"Do this on my behalf" automation is different than "behave with these parameters" autonomy.
There is a grey area, but the differences are who is making decisions, who is being represented, and what is the motivation.
Is it though? The "on my behalf" implies that there are parameters specified.
The difference I'm trying to establish is between directing specific actions (post this at 3:45 Tuesday) and directing behavior (every hour upvote all posts positively portraying hot air balloons)
Digging too far into the difference between automated and autonomous obscures the larger point I was trying to make.
Presumably if an automated account isn't trying to look human then it's unlikely to be mistaken for one, although if it's just there to bump up a number, it might not have to try very hard.
For example, a National Weather Service severe weather social media bot just posting severe weather warnings probably isn't lying to you.
If a Twitter bot named FSB Social Media Account #5527 posted commentary about American politics, there would be very few problems regardless of content.
And now the same thing is slowly happening to bot. It's not only people 'trolling', but it's being done by 'bots'. I believe this is largely a product of the increasing antagonism and radicalism in different 'sects' of our country today. By referring to people you disagree with as 'bots' and 'trolls' it works exactly like wartime propaganda - it dehumanizes and delegitimizes the 'enemy'.
https://www.thisamericanlife.org/545/if-you-dont-have-anythi...
It is like MAD - the assumptions are based upon rational actors which is why they are terrified of proliferation beyond just others getting in their clubhouse.
Unfortunately, the term "NPC" seems to be just a childish insult more often than not. I guess the word "shill" is typically an insult and rarely ever used to actually attack an actual bot or paid-marketer (even though that's the implication).
Troll is the most accurate word I think. I think people are generally happy when a paid employee interacts with the audience (technically a "shill" but... in the case of MMORPGs and such, people like feeling that the game-masters are listening to them).
I'm personally going to stick with the word "troll". I don't really care if they're paid or not, or foreign or not. If they're hampering the discussion, they're a troll. If they're helping the discussion, then they are welcome. So "Troll" is the best distinction, and the main one that matters.
(I'm not a fan, because I feel 'bot' should refer to, well, actual bots, and the better term would be 'shill' or 'sockpuppet', but whatever)
NPC, on the other hand, seems to be used mostly by right-wing actors in reference to left-wing activity/interference.
Personally I'm inclined to think it's largely projection on the side of the NPC-memers/(alt-)right, but I'm sure there's still a bunch of truth to it. Everyone with half a brain uses 'bots' these days.
Of course it matters. A bot can scale to appear as millions of people. So whoever wields the most compelling bots gets a massive force-multiplier for their opinions and objectives. Or more likely, grants that power to the highest bidder.
Meanwhile, labor in Asia is quite cheap ...
Also, they might use bots for the "easy" people, and real human workers for more alert people.
The bots might be a blessing, in some ways, if they result in more people looking for good explanations rather than for assertions or herd direction; assessing ideas rather than their sources.
Normally, without the puppet/bot amplification, the latter guy could be easily ignored. Now his bots can drown out real voices, and that's a problem.
See: the thousands of letters to the FCC regarding net neutrality that were sent by "bots" impersonating real people. Letters written by actual voters were dwarfed by the fake letters.
This quantitative change breeds qualitative change. Any time computers are used against us, be it in information warfare or physical, the reduced cost and decreased risk mean the activation energy is lower and the scale so much greater. The barrier to entry becomes lower, too.
2. What's an acceptable false negative rate before the browser extension is politically problematic
Also, exposing bots on Twitter is how you get the eye of Sauron to notice you. Twitter portrays bots as real users to investors and their advertising clients.