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.

Google Co-Scientist AI fed previous paper with the answer in it

200 pointsby pcfwik3 months ago

6 comments

gukoff3 months ago
&gt; Prof Penadés&#x27; said the tool had in fact done more than successfully replicating his research.<p>&gt; &quot;It&#x27;s not just that the top hypothesis they provide was the right one,&quot; he said.<p>&gt; &quot;It&#x27;s that they provide another four, and all of them made sense.<p>&gt; &quot;And for one of them, we never thought about it, and we&#x27;re now working on that.&quot;<p>The Google&#x27;s co-scientist still seems to make a useful assistant.
hirenj3 months ago
Colour me unsurprised - even not knowing data leakage had occurred, the hypothesis was underwhelming, as I mentioned in a comment on an earlier discussion. I sometimes despair for the state of thinking in science these days given how quickly people fawn over entirely pedestrian thinking and work.<p><a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43105759">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=43105759</a>
euroderf3 months ago
If the AI found a needle in a haystack, that isn&#x27;t bad, is it?
评论 #43165408 未加载
评论 #43166828 未加载
评论 #43165082 未加载
rideontime3 months ago
Instant subscribe when I saw David Gerard&#x27;s name. He cut through the BS in crypto and we need more people like him focusing on AI fraud like this.
评论 #43175096 未加载
mandevil3 months ago
I mean, to be fair, &quot;knowing all of the relevant literature&quot; is a great first step for solving a problem! And a LLM can probably be better about not forgetting than humans are (though I&#x27;m guessing they do much worse on the hallucination front than humans do- humans tend to know when they are playing a hunch). But &quot;This paper from a lower-tier journal two years ago suggests you should look at X&quot; is a very valuable thing to have!
dmitrygr3 months ago
not surprising. why would you expect a simple next token predictor to think?
评论 #43163722 未加载
评论 #43165008 未加载