Self-Consuming Generative Models Go MAD · HackerPapers discussions

Self-Consuming Generative Models Go MAD — mentioned by wavewrangler

Wed, 30 Sep 2026 05:52:32 GMT

Geoffrey Hinton thinks AI is conscious. I know he's extremely accomplished and all, but as it pertains to AI as it exists now we're miles apart (not to say that he's not brilliant, but by difference of opinion). I also don't want to rehash the whole what-is-consciousness philosophical debate, and actually it's a separate question anyway, something doesn't have to be conscious to be dangerous. So I'll stick to the danger bit. Do you use AI much? Not a trick question. What is your usage like? Reason I ask is recall OpenAI hyping GPT-2 with the same language as they are these new models. Doom…

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

Self-Consuming Generative Models Go MAD — mentioned by pankajdoharey

Thu, 25 Jul 2024 05:01:23 GMT

This is a similar problem to what was observed in Diffusion models going "MAD" when trained on synthetic data. https://arxiv.org/abs/2307.01850 . Therefore, going forward AI companies will find it increasingly difficult to get their data by scraping the. web, because web will be full of synthetically generated data.

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

Self-Consuming Generative Models Go MAD — mentioned by SuchAnonMuchWow

Fri, 05 Apr 2024 12:09:26 GMT

MAD: Model Autophagy Disorder (https://arxiv.org/abs/2307.01850 for the paper that introduced the term)

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

Self-Consuming Generative Models Go MAD — HN submission

Thu, 18 Jan 2024 07:01:14 GMT

Self-Consuming Generative Models Go MAD

49 points; 52 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Self-Consuming Generative Models Go MAD — mentioned by matteoraso

Tue, 10 Oct 2023 21:40:29 GMT

>And now that AI generated content is everywhere, that means it's in the next training dataset. Except you can't train AI on AI generated content (https://arxiv.org/pdf/2307.01850.pdf). And no, you can't reliably detect AI generated content either (https://twitter.com/AtoosaChegini/status/1709738910165635364). This is the big one. Every single source of data post March 2023 is corrupted beyond repair, and it's only going to get worse.

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

Self-Consuming Generative Models Go MAD — HN submission

Sun, 06 Aug 2023 11:56:24 GMT

Self-Consuming Generative Models Go Mad

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

Self-Consuming Generative Models Go MAD — mentioned by claytonwramsey

Sun, 16 Jul 2023 18:48:48 GMT

Here's the actual paper that was summarized in the OP: https://arxiv.org/pdf/2307.01850.pdf The paper is quite approachable and worth a quick scan. In essence, reusing output from previous generations results in rather wonky results which are clearly separable from the original training data.

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

Self-Consuming Generative Models Go MAD — mentioned by ImprobableTruth

Sun, 16 Jul 2023 18:46:59 GMT

Best to just hear it straight from the horse's mouth. https://arxiv.org/abs/2307.01850

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