this post was submitted on 26 Jan 2025
212 points (95.3% liked)

Memes

49990 readers
1 users here now

Rules:

  1. Be civil and nice.
  2. Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.

founded 6 years ago
MODERATORS
 
top 50 comments
sorted by: hot top controversial new old
[–] SplashJackson@lemmy.ca 17 points 1 year ago (2 children)

What's a deepseek? Sounds like a search engine?

[–] Karcinogen@discuss.tchncs.de 27 points 1 year ago (6 children)

Deepseek is a Chinese AI company that released Deepseek R1, a direct competitor to ChatGPT.

[–] yogthos@lemmy.ml 27 points 1 year ago (22 children)

You forgot to mention that it's open source.

load more comments (22 replies)
[–] SplashJackson@lemmy.ca 7 points 1 year ago (2 children)

Nice! What are they competing for? I'm new to this AI business thing.

[–] MajorSauce@sh.itjust.works 23 points 1 year ago

So far, they are training models extremely efficiently while having US gatekeeping their GPUs and doing everything they can to slow their progress. Any innovation in having efficient models to operate and train is great for accessibility of the technology and to reduce the environment impacts of this (so far) very wasteful tech.

[–] mukt@lemmy.ml 4 points 1 year ago

Market share, in a speculated market to be in the future.

load more comments (4 replies)
[–] ICastFist@programming.dev 14 points 1 year ago (1 children)

Come on, OP, Altman is still a billionaire. If he got out of the game right now, with OpenAi still unprofitable, he'd still have enough wealth for a dozen generations.

[–] yogthos@lemmy.ml 12 points 1 year ago (1 children)

He's a billionaire based on the valuation of OpenAI, if the company fizzles so does his wealth.

[–] Grapho@lemmy.ml 12 points 1 year ago

🙏🏾🙏🏾🙏🏾

[–] uberstar@lemmy.ml 8 points 1 year ago

I tried DeepSeek, and immediately fell in love.. My only nitpick is that images have to have text on them, otherwise it complains, but for the price of free, I'm basically just asking for too much. Contemporaries be damned.

[–] geneva_convenience@lemmy.ml 3 points 1 year ago

The bubble must be repaired. Pump more cash in!

[–] Sabre363@sh.itjust.works 2 points 1 year ago (1 children)

We doing paid promotions or something on Lemmy now? You sure seem to be pushing this DeepSeek thing pretty hard, op.

[–] yogthos@lemmy.ml 12 points 1 year ago (1 children)

That's right I'm a huge open source shill.

[–] Sabre363@sh.itjust.works 2 points 1 year ago (2 children)

None of this has anything to do with the model being open source or not, plenty of other people have already disputed that claim.

[–] Grapho@lemmy.ml 13 points 1 year ago

It's a model that outperforms the other ones in a bunch of areas with a smaller footprint and which was trained for less than a twentieth of the price, and then it was released as open source.

If it were European or US made nobody would deem it suspicious if somebody talked about it all month, but it's a Chinese breakthrough and god forbid you talk about it for three days

[–] yogthos@lemmy.ml 7 points 1 year ago (9 children)

It has everything to do with the tech being open. You can dispute it all you like, but the fact is that all the code and research behind it is open. Anybody could build a new model from scratch using open data if they wanted to. That's what matters.

load more comments (9 replies)
load more comments
view more: next ›