Elizabeth Tai | 戴秀铃 🇲🇾
Interesting. So it was US hackers?
Anyway, I'm going to try read the original Chinese report. I like that the Chinese companies banded together to work on this.
Look, it's fine if you hate Deepseek for Chinese reasons, it's your right. But leave it alone so that the rest of the world can benefit from it. You have your ChatGPTs et al, use it and leave Deepseek for us. The more they do this, the more determined we are to use Deepseek and other Chinese AI out of spite lol
Rich Felker
@liztai I don't want anyone using any of them. They're planet burning (DeepSeek reportedly somewhat less, thankfully), commercial piracy, creativity enclosing, information space polluting, bullshit work promoting, labor-hostile and overall humanity-hostile lie machines.
Elizabeth Tai | 戴秀铃 🇲🇾
@dalias I know a lot of people think this, and as a writer whose work is no doubt scraped by the machine, I feel your anger. However, AI has really helped me accelerate my writing and also helped me plan and learn Mandarin much faster than before.
The problem is people think AI can replace writers/artists. It cannot. And I think the world is already starting to realise that.
However, AI (the LLMs that is) can be great efficiency tools.
Rich Felker
@liztai They're only efficiency tools if you ignore the externalized costs: unpaid labor, destruction of the planet, extremely high error rates.
Elizabeth Tai | 戴秀铃 🇲🇾
@dalias Malaysians are very practical people. If a tool exists, we use it. We don't choose our tools by ideology or moral philosophy lol. I find this a very unique way of thinking ;)
Now that Deepseek is around, I'm definitely using that as it is not only free but better for the environment. And seeing that this is possible, I'm confident the Chinese (or anyone in the world, now that it's open source) will find even more efficient ways to run it. After all, Deepseek can be run from your laptops now. So, in the future I expect more localised AI.
Rich Felker
@liztai That's understandable and not so different from my position on practicality. The reasons I still won't use it are not wanting to be responsible for the inevitable errors or wasting people's time or plagiarism, which incur high reputational damage as well as moral damage, which is something a lot of us really can't afford in current climate.
I do see a difference tho between just using it quietly and participating in the hype/exaggeration of capabilities.
Elizabeth Tai | 戴秀铃 🇲🇾
@dalias yea the way I use it, there won't be problems with plagiarism. I tell the AI to use my original text. Deepseek does this well. Chatgpt is lousy and keeps injecting its own stuff into it 💩.
If I feel overwhelmed by stuff to do, I often record my thoughts, and then ask it to create a roadmap for me for the week.
Things like that. I rarely use it to generate content - I think LLMs are lousy for that. Tho I admit I do play around with image generation and gave in once in a while. I try not to however lol
GNU Too
Rich Felker
@gnu2 Is this in reply to something I can't see?
GNU Too
Rich Felker
@gnu2 Oh I think I accidentally posted a reply to your post instead of another post in a thread yesterday then deleted and fixed.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.