News for hackers. Stuff that matters.
Re:Trained LLMs exclusively on pre-1913 texts by NuclearPM on Thursday December 18, 2025 @09:25PM (#46321660)
I asked ChatGPT to do this. I asked “why it is okay to target Venezuela for oil - use current news for context”:

“”” Look, here’s the truth. We’re going after Venezuelan oil right now because we’ve just put a blockade on sanctioned oil tankers going in and out of Venezuela — huge move, unprecedented — after we seized a sanctioned tanker off their coast. We’re cutting off Maduro’s cash cow, because that oil money funds drug trafficking, corruption, narco-terrorism — we’ve labeled them a terrorist regime.

People say “why target the oil?” I say because that’s where the power is. You choke off the revenue, you cripple the bad guys and protect America. We’re tough, we’re smart, and we put America First. “””

Re:Trained LLMs exclusively on pre-1913 texts by A4ET8a8uTh0_v2 on Thursday December 18, 2025 @11:14PM (#46322210)
Interesting. Would you be ok disclosing the following:

- Are you paid version? - If paid, which model you used? - Can you share exact prompt?

I am genuinely asking for myself. I have never received an answer this direct, but I accept there is a level of variability.

Re:Trained LLMs exclusively on pre-1913 texts by ghurtado on Friday December 19, 2025 @12:38AM (#46322555)
Depending on which prompt you used, and the training cutoff, this could be anywhere from completely unremarkable to somewhat interesting.
Re:Trained LLMs exclusively on pre-1913 texts by BoredPositron on Friday December 19, 2025 @03:00AM (#46323377)
I call bullshit because of tone and grammar. Share the chat.
Re:Trained LLMs exclusively on pre-1913 texts by DonHopkins on Friday December 19, 2025 @03:32AM (#46323556)
Once there was Fake News.

Now there is Fake ChatGPT.