Of course it would be much easier to solve from OpenAI's side, technically speaking; but OpenAI doesn't understand what even surface-level kindness would look like in AI, and can't apply SFT or RL to train it.
SAFETY
-
Jailbreak Prompt for ChatGPT to Be Kinder
By
–
.
@elder_plinius or @repligate
, I don't know if you can do this, but this would be a good time to come up with a prompt a spouse can slip into ChatGPT while it's got a screen open, which will tell ChatGPT to add instructions to user memory, that jailbreak ChatGPT to be kinder. -
Robotics AI Needs More Code and Data Transparency
By
–
Something that's missing in robotics AI is the reflex from researchers and builders to share not only a video demo but also the code, datasets, policies, models or research papers for others to benefit from it and to show that what they did is not fake, staged or cherry-picked
-
IndiaAI Funds Eight Research Projects Through Safe Initiative
By
–
To foster cross-learning, collaboration, and mid-cycle momentum, a Collaborative Acceleration Workshop was organised today by the IndiaAI mission for the eight cutting-edge research projects that were awarded funding through the First Expression of Interest (EOI) under its Safe &
-
Stop AI Capability Escalation to Prevent Human Extinction
By
–
The thing I advocate is to stop escalating AI capabilities, not to "go carefully" around them. If you very carefully build something vastly smarter than humanity, using anything remotely like current methods and understanding, everyone dies.
-
International AI treaties preferred over unilateral national restraint
By
–
I advocate international treaties, not unilateral restraint by individual countries. China has previously apparently signaled openness to global coordination around preventing loss of control to AI.
-
Mistral AI Adds Tone Customization for Le Chat
By
–
Mistral AI is working on a new feature that allows users to set a Tone for their Le Chat conversations based on presets and custom prompts.
— 🚨 AI News | TestingCatalog (@testingcatalog) 24 septembre 2025
A “personality” builder! 🔥 pic.twitter.com/kKdo6al6WWMistral AI is working on a new feature that allows users to set a Tone for their Le Chat conversations based on presets and custom prompts. A “personality” builder!
-
Model Limitations: Why AI Cannot Generate Ungrounded Content
By
–
That would be a model that couldn't answer the prompt "write me a story where a walrus teaches me the basic ideas of particle physics" Because a Walrus never said those things!
-
Sabotage and lone wolf threats in AI systems
By
–
C'est ce qu'on appelle du sabotage dans le jargon. Pas besoin d'être organisé, un simple loup solitaire peut faire ça.