Most active commenters
  • hereme888(5)
  • bobro(3)

←back to thread

DeepSeek-v3.1

(api-docs.deepseek.com)
776 points wertyk | 18 comments | | HN request time: 0.417s | source | bottom
1. hereme888 ◴[] No.44986048[source]
Reminder DeepSeek is a Chinese company whose headstart is attributed to stealing IP from American companies. Without the huge theft, they'd be nowhere.
replies(8): >>44986091 #>>44986132 #>>44986152 #>>44986163 #>>44986173 #>>44986207 #>>44986219 #>>44986995 #
2. bobro ◴[] No.44986091[source]
Can you contrast this with Western companies? What are the Chinese companies stealing that Western companies aren’t? Do you mean tech or content?
replies(1): >>44986210 #
3. replete ◴[] No.44986132[source]
Reminder that OpenAI is an American company whose headstart is attributed to stealing copyrighted material from everyone else. Without the huge theft, they'd be nowhere.
replies(1): >>44986186 #
4. pphysch ◴[] No.44986152[source]
If an American company did this, it would be "innovative bootstrapping". Yawn.
5. andreashaerter ◴[] No.44986163[source]
I can't say whether those claims are true. But even if they were, it feels selective. Every major AI company trained on oceans of data they didn't create or own. The whole field was built on "borrowing" IP, open-source code, academic papers, datasets, art, text, you name it.

Drawing the line only now... saying this is where copying stops being okay doesn't seem very fair. No AI company is really in a position to whine about it from my POV (ignoring any lawyer POV). Cue the world's smallest violin

6. computerex ◴[] No.44986173[source]
I find it hilarious you felt the need to make this comment in defense of American LLMs. You know that American LLMs aren’t trained ethically either, right? Many people’s data was used for training without their permission.

BTW DeepSeek has contributed a lot, with actual white papers describing in detail their optimizations. How are the rest of the American AI labs doing in contributing research and helping one another advance the field?

7. hereme888 ◴[] No.44986186[source]
Last I checked, as it concerns the training of their models, all legal challenges are pending. No theft has yet been proven, as they used publicly available data.
replies(3): >>44986292 #>>44986479 #>>44990497 #
8. pdntspa ◴[] No.44986207[source]
As if those american companies played fair with training their AIs

It's theft all the way down, son

9. hereme888 ◴[] No.44986210[source]
Ethics of Chinese vs. Western companies? Everything. I'm sure you're aware of how many hundreds of $billions of American IP are stolen by Chinese companies.
replies(1): >>44986353 #
10. nancyminusone ◴[] No.44986219[source]
Most ironic comment I've yet laid eyes on.
11. dghlsakjg ◴[] No.44986292{3}[source]
Legal != ethical
12. bobro ◴[] No.44986353{3}[source]
I’m not asking broadly about difference in ethics. I’m asking specifically about IP theft in the AI space.
replies(1): >>44986798 #
13. torginus ◴[] No.44986479{3}[source]
In contrast to your legally watertight accusations.
14. hereme888 ◴[] No.44986798{4}[source]
I'm not aware of any proven IP theft by American companies in the AI space. Many pending legal challenges. None yet proven.
replies(1): >>44988716 #
15. mach5 ◴[] No.44986995[source]
who cares
16. bobro ◴[] No.44988716{5}[source]
Alright. So there’s proven IP theft by the Chinese companies with completed legal proceedings?
replies(1): >>44991826 #
17. skinnymuch ◴[] No.44990497{3}[source]
Do you think the whole world follows America's legal system? Western and American exceptionalists...
18. hereme888 ◴[] No.44991826{6}[source]
With completed legal proceedings, at least two cases: Xiaolang Zhang was sentenced in 2024 for stealing Apple's AI autonomous vehicle tech for Chinese AI company XPeng. Xiang Haitao was sentenced in 2022 for stealing Monsanto's AI predictive algorithm for a Chinese research institute.