Thread

1. hyperionultra · 1790171488 · focus · HN ↗

   Well, since llms need matterial for self-education, someone still had to be inteligent enough. Or llms will freeze their “intelect” at some point.

   1. BenzeneDream · 1790172041 · focus · HN ↗

      Why would it freeze if they can create more training material. Its already been shown synthetic training data is just as good if not better.

      1. timcobb · 1790172619 · focus · HN ↗

         > Its already been shown synthetic training data is just as good if not better.

         Source, please! I'm a machine learning layperson but how can this not be overfitting?

         1. esseph · 1790173966 · focus · HN ↗

            <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2609.26457v1" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2609.26457v1

            &gt; AIDE demonstrates that recursive self-improvement at the harness layer can produce transferable gains in an AI research agent’s research efficiency. During the recursive self-improvement run, the loop accepted seven rewrites, each under a fixed evaluation budget (section 3.2). Under this fixed evaluation budget, gains in optimization capability on AI R&amp;D tasks translate to gains in research efficiency.&quot;

            <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2609.11873" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2609.11873

            &gt; The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement

Unofficial Hacker News client; not affiliated with Y Combinator.