2 ms·
I wonder if the current revolution in AI will accidentally kill online message boards. Here's the line of reasoning. Some of these steps are probably already o
by lordnacho 10y ago
I wonder if the current revolution in AI will accidentally kill online message boards.
Here's the line of reasoning. Some of these steps are probably already occurring.
- It will be easy write a bot that understands what's being said.
- A bot can follow many threads at once, multiplying the reading workload for humans.
- A bot can pull large amounts of information together into a cogent argument, or at least one that cannot be dismissed by a human without a fair few minutes of reading. For instance, if I want to troll people by coming with reasons why women should not be allowed to vote, my bot could spin a yarn about how Switzerland is an advanced industrial country that only in 1991 allowed its remaining women to vote. I could pull data about women being left-leaning, more emotional, and so on. And the data needn't actually be true, I can quote sources with wrong numbers, create new sources, and so on. You, the interested debater, would have to follow this pile of crap and debunk each item, eating up your whole evening.
- In order to defeat such a bot, someone will invent a fact-checking bot (Like a GAN, it could decide whether you were botting your argumentation). Sounds good, but wait. A fact-checking bot would have to present its side of the story. And how would it do that? Just like a human, here's your points, here's why they're wrong, here's the actual facts, and so on.
- So now every message board is bots talking to bots.
- Humans will have to fall back to heuristics, like "read the first line and decide if there's a point". But of course a bot can figure out what the weaknesses in your heuristics are.
- People will get sick of writing on boards, because well, they're not for people.