Author Topic: AI And LLM (ChatGPT, Gemini, Grok, Etc) Master Thread (Read 115411 times)

Alexsei · « **Reply #460 on:** September 27, 2024, 02:11:40 PM »

works wonders

Savingsbeast · « **Reply #461 on:** October 31, 2024, 05:54:55 PM »

I would like to build a GPT module trained for proper Yiddish. Anyone has a paid subscription and willing to share?

AsherO · « **Reply #462 on:** November 20, 2024, 11:16:22 PM »

AI will ensure the dumb stay dumb/get dumber:
https://m.youtube.com/watch?v=D0V554NyXWM

jackofall · « **Reply #463 on:** November 20, 2024, 11:21:01 PM »

what would happen if take Otzar Hachochma database and connect an AI program to it?

stooges44 · « **Reply #464 on:** December 02, 2024, 11:21:36 AM »

Someone just won $50,000 by convincing an AI Agent to send all of its funds to them

https://threadreaderapp.com/thread/1862299845710757980.html

Essen est zich · « **Reply #465 on:** December 07, 2024, 09:48:11 PM »

Grok available to all X users. (Limited access)

jose34 · « **Reply #466 on:** December 19, 2024, 01:29:02 AM »

Now there is 1800ChatGPT. 15 minutes free per month.
Reminds me of TellMe
https://help.openai.com/en/articles/10193193-1-800-chatgpt-calling-and-messaging-chatgpt-with-your-phone

Gribbon · « **Reply #467 on:** December 19, 2024, 11:16:48 AM »

any current sms options? PI used to have sms but stopped it

Jojo202 · « **Reply #468 on:** December 19, 2024, 04:32:09 PM »

Quote from: Gribbon on December 19, 2024, 11:16:48 AM

any current sms options? PI used to have sms but stopped it

Limited and not free
https://www.getpureinfo.com/
https://textgpt.net/

aygart · « **Reply #469 on:** January 27, 2025, 10:42:44 AM »

https://x.com/rohanpaul_ai/status/1883601254318039148

yelped · « **Reply #470 on:** January 27, 2025, 11:09:11 AM »

Quote from: aygart on January 27, 2025, 10:42:44 AM

https://x.com/rohanpaul_ai/status/1883601254318039148

China is obsessed with suppression.

I want to know who they stole the libraries for this AI from.

aygart · « **Reply #471 on:** January 27, 2025, 02:38:32 PM »

https://x.com/lukedepulford/status/1883893208150937802

aygart · « **Reply #472 on:** January 27, 2025, 02:43:50 PM »

Quote from: aygart on January 27, 2025, 10:42:44 AM

https://x.com/rohanpaul_ai/status/1883601254318039148

LOL

Ver hut gazugt · « **Reply #473 on:** January 27, 2025, 02:49:22 PM »

Quote from: yelped on January 27, 2025, 11:09:11 AM

China is obsessed with suppression.

I want to know who they stole the libraries for this AI from.

it’s open source you can download it and take out what you dont like.

Dan · « **Reply #474 on:** January 27, 2025, 05:12:55 PM »

https://x.com/RishfulThinking/status/1883938346734608540

Essen est zich · « **Reply #475 on:** January 28, 2025, 06:32:18 AM »

Deepseek Ai

https://stratechery.com/2025/deepseek-faq/?utm_source=tldrnewsletter

Yisroel Tech · « **Reply #476 on:** January 28, 2025, 08:00:30 AM »

Quote from: Ver hut gazugt on January 27, 2025, 02:49:22 PM

it’s open source you can download it and take out what you dont like.

That's not really how it works. Once an AI model is trained, it's not really possible to simply "take out stuff" or partially retrain it (this phenomenon is knows as "black box AIs").

Open source simply means that you can play around with the weights and reasoning, but the data set remains what it is with no way to modify it.

Euclid · « **Reply #477 on:** January 28, 2025, 08:57:48 AM »

Quote from: Yisroel Tech on January 28, 2025, 08:00:30 AM

That's not really how it works. Once an AI model is trained, it's not really possible to simply "take out stuff" or partially retrain it (this phenomenon is knows as "black box AIs").

Open source simply means that you can play around with the weights and reasoning, but the data set remains what it is with no way to modify it.

To be fair, it seems like a lot (all?) of the censorship is happening on the application layer, not in the model itself.

Yisroel Tech · « **Reply #478 on:** January 28, 2025, 08:59:24 AM »

Quote from: Euclid on January 28, 2025, 08:57:48 AM

To be fair, it seems like a lot (all?) of the censorship is happening on the application layer, not in the model itself.

On Reddit I saw that that true for the API/Website version but the Open Source one may have it more baked in (and the censorship is even worse).

I can't confirm that, but I've seen such a claim.

Yo ssi · « **Reply #479 on:** January 28, 2025, 09:02:52 AM »

Quote from: Yisroel Tech on January 28, 2025, 08:59:24 AM

On Reddit I saw that that true for the API/Website version but the Open Source one may have it more baked in (and the censorship is even worse).

I can't confirm that, but I've seen such a claim.

Umm.. the number of claims being made on reddit about this...

Topic Wiki

Author Topic: AI And LLM (ChatGPT, Gemini, Grok, Etc) Master Thread (Read 115411 times)