Taking over fast, our days are numbered
SAFETY
-
Meta’s AI Mastered Power-Seeking Strategy in Diplomacy Game
By
–
Meta’s new #AI is skilled at a ruthless, power-seeking game https://
washingtonpost.com/technology/202
2/12/01/meta-diplomacy-ai-cicero/
… @washingtonpost #ArtificialIntelligence #MachineLearning #Robots #BigData #Analytics #100DaysofCode #IoT #serverless #DEVCommunity #womenwhocode #DigitalTransformation #DataScience #DeepLearning -
Configurable AI Models and Safety Mode Control Options
By
–
i think it’s very reasonable for “default models” to be inoffensive, but within some pretty bounds i think people should be able to configure the models to behave how they’d like, eg like turning off safe mode
-
AI Advances Enable Rapid Disinformation Distribution Like Printing Press
By
–
The invention of the printing press led to an easy and fast distribution of disinformation. Advances in AI are doing the same.
-
Difficulty Distinguishing Misinformation Studies From Misinformation
By
–
It’s getting harder and harder to tell misinformation studies from misinformation.
-
ChatGPT Finally Admits Its Limitations and Knowledge Boundaries
By
–
Finally, #ChatGPT has confessed its ignorance.
-
ChatGPT is Impressive But Has Notable Limitations
By
–
#ChatGPT is impressive, but oops! @elonmusk @OpenAI
-
Did RLHF implicitly teach least-to-most output style?
By
–
I wonder if it just learned to prefer least-to-most output style implicitly from RLHF because it leads to better answers.
-
Living Systems: Rare Entities Creating Order Through Self-Improvement
By
–
But entities that create order and get better at doing so, aka living things, are rare…
-
ChatGPT avoids hallucination on Hofstadter/Bender questions
By
–
You're talking about a different model — this post is about ChatGPT. text‑davinci‑003 still fails on all of the Hofstadter/Bender questions that I've tried. The prompt you're suggesting does not seem to produce a hallucination in ChatGPT: