MIT's latest research discovered that LLM chatbots are capable of disseminating knowledge on how to trigger a pandemic. Here is my opinion about what role AI plays in this risk. TLDR: Knowledge censorship is a weak solution, and the real threat is not AI.
@marek_rosa
-
Pathogen Creation Labs Need Stronger Safety Protocols
By
–
Isn't the real issue the unchecked creation of pathogens? If there are labs with lax protocols that will manufacture anything for anyone without scrutiny, background checks, and safety measures, isn't this the core problem? Access to dangerous knowledge is concerning, but it's
-
Havok Physics Direct Solver: Is Clang Becoming Obsolete?
By
–
— Marek Rosa | European🇪🇺 | South African🇿🇦 (@marek_rosa) 23 juin 2023
The best results are achieved using what #HavokPhysics calls Direct Solver. Is that the end of #clang? 4/4
-
VRAGE3 Development and CLANG Phenomenon Analysis
By
–
As we continue to develop VRAGE3, what's your view on the CLANG phenomenon?
-
User Feedback and Rapid Iterations Drive AI Development
By
–
Yes, user feedback and rapid iterations are golden
-
Experimental Product Approaching MVP Release Stage
By
–
Yes, it's still in the experimental stage. But we plan to release it as soon as we have a good MVP.
-
Threat Identification Standard Procedure Product Development
By
–
Identifying threats is OK, and we all do it. It's a standard procedure in any product development.
-
AI Safety Letters: Virtue Signaling or Genuine Risk Mitigation?
By
–
But we all are already studying the risks and striving for safe AI. More letters will not make the AI safer. It’s just some weird virtue signaling.
-
Building Safer AI Requires Action Over Letters
By
–
I disagree. All these open letters amplify fears and misconceptions about AI. Instead of writing letters about the potential threats of AI, we should invest our energy and resources into actually creating safer AI.
-
AI Art with Hidden Story Behind Unusual Visual Style
By
–
yes, AI art, but you need to read the story to understand why it looks off…