THE VIDEO SHELF · gpt-4chan experiment
GPT-4chan: This is the worst AI ever
Yannic Kilcher explains how he trained a language model on three years of 4chan's /pol/ board and let it post on the site. He evaluates how good the model was and discusses the ethics and fallout.
Why HN readers recommended it
HN points to it as the definitive account of the GPT-4chan experiment, used in debates on LLM bias, truthfulness and bots.
This is a summary from the source dataset, not an endorsement by Hacker News or the video creator.
Video & recommendation details
- Creator / channel
- Yannic Kilcher
- Topic
- gpt-4chan experiment
- Format
- commentary
- Level
- general
- HN recommenders in dataset
- 3
- HN mentions in dataset
- 18
- Recommendation years in dataset
- 2022–2026
These counts and years describe the included Hacker News records; they are not YouTube views or the video’s publication date.
From the original HN conversations
Excerpts from the recorded comments. Follow each source for the full comment and surrounding discussion.
In my opinion the granddaddy of all 4chan CAPTCHA busts is still Yannick Kilcher’s GPT-J tune on “Raiders of the Lost Kek” set, and might be the coolest thing an LLM has ever done on video: [VIDEO]
benreesman on Hacker News · 2024-11
In Breaking the 4Chan CAPTCHAThere is a solid video explanation (though the author is undermining the toxicity of the average 4chan /pol/ posts): [VIDEO] . A really interesting experiment.
thegreatdukd on Hacker News · 2022-06
In GPT-4chanI still love this experiment: [VIDEO] The theories about the origins of these commenters were amazing. There's nothing to say that other people are not doing the exact same thing right now here on HN and other places. And very few people would even notice, …
sigmoid10 on Hacker News · 2024-03
In Reddit Launches Long-Awaited IPO with $748M Target
Summaries, classifications and recommendation records are from the HackerTube source dataset. Read about this library and its limitations.