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.

A path to O1 open source

133 pointsby bchelli4 months ago

16 comments

mrayycombi4 months ago
From the first few paragraphs it doesn&#x27;t pass the sniff test for me.<p>&quot;Now AI has made everything more complex!&quot; &quot;AI is embedded in everything we do&quot;...<p>Sounds like marketing gibberish and obfuscation, combined with self promotion.<p>That&#x27;s just my read at first sniff.
评论 #42591440 未加载
kcorbitt4 months ago
Lots of folks working on open-source reasoning models trained with reinforcement learning right now. The best one atm appears to be Alibaba&#x27;s 32B-parameter QwQ: <a href="https:&#x2F;&#x2F;qwenlm.github.io&#x2F;blog&#x2F;qwq-32b-preview&#x2F;" rel="nofollow">https:&#x2F;&#x2F;qwenlm.github.io&#x2F;blog&#x2F;qwq-32b-preview&#x2F;</a><p>I also recently wrote a blog explaining how reinforcement fine-tuning works, which is likely at least part of the pipeline used to train o1: <a href="https:&#x2F;&#x2F;openpipe.ai&#x2F;blog&#x2F;openai-rft">https:&#x2F;&#x2F;openpipe.ai&#x2F;blog&#x2F;openai-rft</a>
评论 #42592133 未加载
xoxosc4 months ago
Is it me or the first link on the site <a href="http:&#x2F;&#x2F;ability.openai" rel="nofollow">http:&#x2F;&#x2F;ability.openai</a> is broken? Is openai a tld now?
评论 #42590749 未加载
评论 #42590891 未加载
评论 #42601445 未加载
评论 #42590712 未加载
rrr_oh_man4 months ago
I don’t think anyone who upvoted this has read more than one sentence of this paper.
评论 #42590876 未加载
评论 #42590979 未加载
评论 #42590880 未加载
jeanlucas4 months ago
Looks like China will lead the next generation of open source tech.
评论 #42590438 未加载
评论 #42590783 未加载
评论 #42590454 未加载
评论 #42591764 未加载
mtkd4 months ago
There was a Berman video on it earlier today that summarises it<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=-haWhgmUheA" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=-haWhgmUheA</a><p>Detail starts about 7mins in
fuddle4 months ago
It&#x27;s ironic that they are attempting to open source a model from &quot;OpenAI&quot;.
评论 #42590760 未加载
评论 #42590750 未加载
评论 #42590564 未加载
评论 #42590626 未加载
评论 #42591345 未加载
halayli4 months ago
&gt; has claimed that the main techinique behinds o1 is the reinforcement learining.<p>Typos in the first sentence of the paper doesn&#x27;t give confidence that I am about to read something worthwhile.
评论 #42591408 未加载
评论 #42591300 未加载
newyankee4 months ago
I guess now the strategy of OpenAI would be to keep the small edge all the time, integrate it with businesses fast &amp; possibly kickstart new businesses by supporting them and trying to be synonymous with the best in AI (may be with Deepmind). I cannot think of any other moat, unless somehow they have a lot of proprietary and useful data (like in company) that others cannot replicate
评论 #42591132 未加载
currymj4 months ago
many people are dismissing this paper because it has errors in spelling and grammar.<p>this is a terrible heuristic for evaluating AI papers. If you use it, you will miss a lot of good work by very strong researchers with below-average English writing skills.<p>I have not read this paper carefully so claim nothing one way or the other about its quality. It superficially seems like a pleasant and timely survey although a little flag-planty.
Oras4 months ago
First line in the abstract<p>&gt; OpenAI o1 represents a significant milestone in Artificial Inteiligence,<p><i>Inteiligence</i><p>Safe to say OpenAI has nothing to worry about
评论 #42590835 未加载
评论 #42590566 未加载
评论 #42590684 未加载
HarHarVeryFunny4 months ago
Sounds about right.<p>How long until we see DeepSeek-o1 ?
评论 #42590563 未加载
punnerud4 months ago
Not one word in the article about using embeddings for the reasoning, a bit strange?
评论 #42590902 未加载
chefandy4 months ago
It seems to me that for the most capable and useful models, openness almost exclusively benefits businesses, or maybe academic organizations with money for serious hardware. I know what I can run on my 4090 at home but the results pale in comparison to the commercial services. I see why people consider these matters important from a theoretical standpoint but from a practical standpoint it doesn’t seem particularly consequential. I self-host a few FOSS server applications that are primarily sold as SaaS subscriptions, and folks are often very critical of those businesses benefitting from the “open source” label because they’re often seemingly deliberately difficult to self-host. This seems to be an order of magnitude less open than that. Is there some use case for people with reasonable hardware that I’m just not aware of?
评论 #42590772 未加载
评论 #42591005 未加载
jbegley4 months ago
lmao at “learining” being misspelled in the first sentence
评论 #42590505 未加载
评论 #42590490 未加载
mig394 months ago
Is this a joke? Why are there so many spelling mistakes? &quot;Inteiligence&quot; &quot;challanging&quot; “learining”<p>What is going on?
评论 #42590572 未加载