Our position on open-weights models

Re:Our position on open-weights models by rob74 on Tuesday July 28, 2026 @04:36AM (#49081063)

It's the machine learning platform that started with "developing a chatbot app targeted at teenagers" (according to https://en.wikipedia.org/wiki/Hugging_Face), which helps explain the weird name (which still makes me think more of one of the development stages of the xenomorphs from Alien than of the emoji, but that's probably just me).

[ Parent | on HN ]

Re:Our position on open-weights models by nyeah on Tuesday July 28, 2026 @08:17AM (#49082675)

It's definitely not just you :)

[ Parent | on HN ]

Re:Our position on open-weights models by smallmancontrov on Tuesday July 28, 2026 @09:37AM (#49083691)

I have it bookmarked as just the emoji because whenever I write the text I still think about the alien spider.

[ Parent | on HN ]

Re:Our position on open-weights models by coldtea on Tuesday July 28, 2026 @05:59AM (#49081643)

If you're not familiar with such overall big players, would anything help short of looking at some 101 first?

[ Parent | on HN ]

Re:Our position on open-weights models by soblemprolver on Tuesday July 28, 2026 @06:38AM (#49081911)

I am familiar, yet used this comment to confirm my initial assumption.

Abbreviations carry unnecessary cost on the side of the readers. There is always the chance that there is another shorthand for something else that is important and so there is a need to double check interpretations.

So thanks for the comment and no, a 101 wouldn't even help with this problem.

[ Parent | on HN ]

Re:Our position on open-weights models by coldtea on Tuesday July 28, 2026 @07:23AM (#49082247)

>Abbreviations carry unnecessary cost on the side of the readers.

They do, but the parent's question wasn't about the abbrebation (for which I'd cut them some slack) but extended to:

"Is that supposed to help? WTF is a HuggingFace?".

>So thanks for the comment and no, a 101 wouldn't even help with this problem.

Wasn't addressed to you though (don't even see any other comment of you in this subthread), but to @markdown, whose issue wasn't with the abbreviation alone.

[ Parent | on HN ]

Re:Our position on open-weights models by tough on Tuesday July 28, 2026 @06:51AM (#49082008)

if you google "hf hack" the first result is OpenAI post-mortem official blog post so...

[ Parent | on HN ]

Re:Our position on open-weights models by witrak on Tuesday July 28, 2026 @07:25AM (#49082263)

Lol! Does it matter? Using an unclear acronym in the first reference to an object, besides adding energy waste (for the execution of unnecessary searches by users unfamiliar with the subject), may only give the author of the comment the thrill of "being well informed, an insider".

[ Parent | on HN ]

Re:Our position on open-weights models by bethekidyouwant on Tuesday July 28, 2026 @08:35AM (#49082884)

Wtf is a coca-cola?

[ Parent | on HN ]

Re:Our position on open-weights models by Barbing on Tuesday July 28, 2026 @11:03AM (#49085032)

Curse of knowledge. / Unsafe assumption.

Some users feel IP address, ISP, and LLM are clear. Fewer of those users will also reference HF. I have had to stop myself and type out Hugging Face.

HR perhaps? https://en.wikipedia.org/wiki/Hanlon%27s_razor

[ Parent | on HN ]

Re:Our position on open-weights models by swiftcoder on Tuesday July 28, 2026 @04:31AM (#49081029)

The biggest platform for the distribution of open-weight models

[ Parent | on HN ]

Re:Our position on open-weights models by gertrunde on Tuesday July 28, 2026 @04:41AM (#49081103)

For context:

https://news.ycombinator.com/item?id=48997548

[ Parent | on HN ]

Re:Our position on open-weights models by tempfile on Tuesday July 28, 2026 @05:08AM (#49081281)

Surely you can just google this.

[ Parent | on HN ]