Has anyone tried DeepSeek yet?

[mai72]

OMG, it is so cool. Its like I'm talking to a real person.

I just found a friend.

[sleeps]

i can talk with it without feeling ashamed of my obsession with muscular shark girls.

[Ele]

Not yet, but I will. Ironically, ChatGPT is too left-leaning. Maybe a Chinese AI would be more data-centered instead of focused on emotional political correctness.

[Heitor Villa-Lobos]

Yes. Its much better than ChatGPT for non-English speakers.

ChatGPT is racist and narrow-minded.

GPT once refused to use the thousands separators as a dot, and decimals as a comma. Even if more countries use it this way (most countries of America, most countries of Europe, etc), it insisted on the US way. I only speak Portuguese to it, yet it defies me on my own language.

I'm not surprised if it also defaults to backwards dates and imperial units as well.

ChatGPT doesn't understand us, and does not respect us.

[GunsForBucks]

Anyone else creeped out?

[Heitor Villa-Lobos]

Originally posted by GunsForBucks:

Anyone else creeped out?

Why??
ChatGPT is literally learning from your inputs. Major data leaks have happened because people were not careful. You are sending everything you type into ChatGPT text box to OpenAI.

DeepSeek, on the other hand, is open source and there is no data collection for spionage or political harassment there. Also, you can run it locally and off-line.

Its like streaming your bathroom to youtube, but being scared to watch videos on BiliBili: prejudice and paranoia.

[GunsForBucks]

Normalization freaks me out.

It's the way of censorship and control.

[BigBallinChester]

Deepseek is CCP spyware that is censored by their government. Not to mention the CCP spent billions to steal OpenAI training data to make.

They also leaked data on every single user interaction with it, accounts,discussion everything. It will also politically harass you.

llama3 is open source and not censored by a government.

[Heitor Villa-Lobos]

^ what a load of lies and misinformation!

OpenAI stole data from us all. They're criminals and the law just bended to fit their practices. This was widely debated worldwide! Artists are still fighting unconsented data collection. Pricacy activists are denouncing the dragnet surveillance being used to train LLMs.

China is a country, not a company.

There has been many other breaches of OpenAI and LLMs, its not unique to Chinese companies.

Llama is owned by Meta, an infamous genocide agent around the world. They should be regulated.

[Rumpelcrutchskin]

Have fun with your new friend Xi.

[Hong Kong Phooey]

Hey, has anyone tried talking with mai72. I'm sick of talking to AI.

[Råb!d]

All AI shows bias. The fact that they hallucinate is disconcerting. The rush to embrace AI for national defense or offense is scary. The replacing of workers is eliminating a lot of entry-level and less-skilled jobs is disturbing. The use of AI to create deep fakes is troublesome.