When Large Language Models are More PersuasiveThan Incentivized Humans, and Why · HackerPapers discussions

When Large Language Models are More PersuasiveThan Incentivized Humans, and Why — mentioned by keeda

Mon, 21 Sep 2026 20:55:00 GMT

Haha, sure! Some of the studies I read about, though admittedly I haven't read most of them in depth: https://www.fbk.eu/en/press-releases/artificial-intelligence... https://www.nature.com/articles/s41467-025-61345-5 https://arxiv.org/abs/2505.09662 https://arxiv.org/html/2411.06837v2 https://arxiv.org/html/2604.22109v2

Paper and discussion history. A mention is not an endorsement.

When Large Language Models are More PersuasiveThan Incentivized Humans, and Why — mentioned by intended

Sun, 01 Jun 2025 13:27:09 GMT

This keeps me up at night too. I’d like to stake the position that LLMs are antagonistic to the (beleaguered) idea of an internet. LLMs increase the burden of effort on users to successfully share information with other humans. LLMs are already close to indistinguishable from humans in chat; Bots are already better at persuading humans[1]. Suggesting that users who feel ineffective at conveying their ideas online, are better served by having a bot do it for them. All of this, is effectively putting a fitness function on online interactions, increasing the cognitive effort required for humans…

Paper and discussion history. A mention is not an endorsement.

When Large Language Models are More PersuasiveThan Incentivized Humans, and Why — HN submission

Sat, 17 May 2025 20:05:09 GMT

LLMs are more persuasive than incentivized human persuaders

140 points; 116 discussion replies on HN. Paper and source history. A submission is not an endorsement.

When Large Language Models are More PersuasiveThan Incentivized Humans, and Why — HN submission

Fri, 16 May 2025 20:34:56 GMT

Large Language Models Are More Persuasive Than Incentivized Human Persuaders

4 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.