TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

The “it” in AI models is the dataset

3 pointsby ziyunlialmost 2 years ago

1 comment

ggmalmost 2 years ago
<i>What that means is not only that they learn what it means to be a dog or a cat,</i><p>NO. Please.. they have no idea &quot;what it is like to be a bat&quot; they have been trained to respond to images with a high enough threshold of dog-ness or cat-ness about them.<p>&quot;what it means to be&quot; is not a phrase anyone in this field should be using.<p><i>Then, when you refer to “Lambda”, “ChatGPT”, “Bard”, or “Claude” then, it’s not the model weights that you are referring to. It’s the dataset.</i><p>This I can get behind. &quot;its statistics. whichever method you use, the underlying model is most probably like the input set its derived from, because the quality it reflects is in that dataset&quot;<p>If however, the divergences between them turned out to be interesting, I&#x27;d say &quot;its not the dataset, its how people intuit fit against the dataset&quot;