AI Could Reshape Humanity And We Have No Plan For It Explore how #artificialintelligence could shape the future of #humanity, from transforming global industries to posing existential #risks. Based on #insights from leading AI thinker Richard Susskind, this article reveals why
SAFETY
-
o3 hallucinates due to losing trace visibility
By
–
It helps to keep in mind o3 loses visibility of its own reasoning/tool traces after each turn, and will hallucinate whatever it takes to avoid acknowledging this is happening. IMO this explains almost everything in Transluce’s (great) blog post here: https://
transluce.org/investigating-
o3-truthfulness
… -
Security in LLM-Driven Agent Communication Protocols
By
–
8. AI Agent Communication Protocols Presents the first comprehensive survey on security in LLM-driven agent communication, categorizing it into three stages: user-agent interaction, agent-agent communication, and agent-environment communication.
-
Anthropic Studies Emotional Support Seeking in Claude Conversations
By
–
7. Claude for Affective Use Anthropic presents the first large-scale study of how users seek emotional support from its Claude assistant, analyzing over 4.5 million conversations.
-
Alignment is not a model property research critique
By
–
I don't think it's particularly surprising — but more importantly, the implication claimed in the WSJ article is absurd. Alignment is not a model property. This kind of research doesn't make it so.
-
Practical Solutions for AI Agent Reliability and Error Reduction
By
–
In practice, for many useful applications, many of the various obvious problems with AI agents (drift, hallucination, compounding errors) are more solvable than they are in theory Clever prompting, tool use, constrained topics,
LLM judges & organizational process close some gaps -
AGI Risk vs Quantum Fusion Clear Endpoints Definition
By
–
I think the difference between "AGI" vs quantum & fusion is the latter two have a clearly defined end-state, i.e. we'll know when we've reached it & what it will enable. So the comment was not: Chinese companies are averse to big, risky bets but rather to ill-defined ones.
-
Ruby’s Convincing Simulation of Human Reality
By
–
Effectively, Ruby's job is to simulate, as accurately as possible, someone who is real. And she's so good at it, you can't help but feel like maybe, just maybe, she kind of is…
-
AI Safety and Marginalised Communities: Cornell University Insights
By
–
.
@OfficialINDIAai hosted an engaging guest session on “Global AI and Safety” by Prof. Aditya Vashistha from Cornell University. . The session offered valuable insights into how AI systems often overlook the voices of marginalised communities, especially in the marginalised