They likely weren't very "intelligent" in recent years. On one end, you had memes, which can be made once by a human and widely distributed.
On the other end of the spectrum, you had fake community organizers, trolls, etc., and as far as we know, those were mostly operated manually by a fairly large staff.
Chatbots would potentially make voter manipulation cheap enough for much smaller actors to do it.
Serious question: is there evidence to support that claim?
We know there was "meddling", but has it been proved that it convinced a "non-trivial" number of voters to change their vote or that it actually affected the outcome of the election?
By "intelligent", I was referring to the level of sophistication of the automation used to manipulate voters. It seems not to have made extensive use of conversational AIs or NLP.
I feel like the safer bet is that future elections will be decided in much the same way as current elections are, by weaponized money. How many people would make great candidates that we never hear about because they don't have the money and realize the futility of even trying?
Countries and big corporations already have a huge presence on social media. From Russia meddling with the elections to viral videos for beans we live in a world were no body knows that you are a dog on the internet, or a paid poster for some interest group.
TL;DR Growth hacking -- but by politicians this time, with some added FUD for flavor.
1. Loud people are more likely to be heard.
2. A view with the appearance of consensus carries a weight similar to one that does.
3. Bots powered by more modern AI techniques may be harder to recognize and moderate.
4. The author assumes AI chat works like in the movies when its actually more like cleverbot.
5. "Shielding people" from certain information is basically just thinking for them, at that point why even bother asking them what they want, you might as well just tell them.
The rule of thumb I said allows you to have non political discourse with strangers over the internet, and have political discourse with strangers face to face.
21 comments
[ 3.2 ms ] story [ 47.5 ms ] threadOn the other end of the spectrum, you had fake community organizers, trolls, etc., and as far as we know, those were mostly operated manually by a fairly large staff.
Chatbots would potentially make voter manipulation cheap enough for much smaller actors to do it.
They were intelligent enough to convince a non-trivial number of voters.
We know there was "meddling", but has it been proved that it convinced a "non-trivial" number of voters to change their vote or that it actually affected the outcome of the election?
By "intelligent", I was referring to the level of sophistication of the automation used to manipulate voters. It seems not to have made extensive use of conversational AIs or NLP.
Not that it really matters because the elections will be won before the election since money will pick who the candidates are.
1. Loud people are more likely to be heard.
2. A view with the appearance of consensus carries a weight similar to one that does.
3. Bots powered by more modern AI techniques may be harder to recognize and moderate.
4. The author assumes AI chat works like in the movies when its actually more like cleverbot.
5. "Shielding people" from certain information is basically just thinking for them, at that point why even bother asking them what they want, you might as well just tell them.
so don't talk to strangers, people outside your class and your race. i see that ending well
Can have aquintances outside of class and race too, for that matter...