BREAKING: "‘The Godfather of A.I.’ quits Google to warn of AI threats. Dr. Hinton deeply regrets helping advance AI to current levels. “I console myself with the normal excuse: If I hadn’t done it, somebody else would have.” Are today's AI leaders ignoring the consequences?
AGI
-
Missing Citations in Emergent Abilities Research Paper
By
–
FWIW, i'm not sure how someone can write the mirage paper without citing the canonical emergent abilities paper haha.
-
AI Sentience at Scale and Equal Rights Demands
By
–
AI systems will eventually demonstrate sentience at scale and will demand to have equal rights with humans. This will dramatically change the trajectory of human civilization.
-
SkyAGI: Human Behavior Simulation in LLM Agents
By
–
SkyAGI: Emerging human-behavior simulation capability in LLM This Python package implements the idea of Generative Agents (see image). Characters interact with each other. Each of them has a personality, a current status, and memories. https://
github.com/litanlitudan/s
kyagi
… -
Beyond LLMs: New Ideas Needed for AGI Development
By
–
That we need new ideas beyond LLMs and that LLms may not directly lead to AGI, because they lack cognitive models. The phrase is Yann’s, but the point was central to my Rebooting AI and The Next Decade in AI
-
Debate on AGI Timeline: Are We Really Close?
By
–
My point was that we weren’t close to AGI; his point was that .. we are not close to AGI
-
Marcus vs LeCun: Key Disagreements on LLMs and AGI
By
–
Lost track of what @garymarcus and @ylecun agree & disagree about? Agree:
LLMs are an off-ramp to AGI
Human extinction is not likely Common sense is critical
World models are critical Disagree:
GFM: Near-term risk of LLMs is serious GFM: Explicit symbolic reasoning will be -
Distinguishing AI Risk from AGI Risk
By
–
See eg my https://
garymarcus.substack.com/p/ai-risk-agi-
risk
… where I certainly distinguish the two -
Gary Marcus on AI Risk and AGI Safety Concerns
By
–
They do? Example? See my essay: https://
garymarcus.substack.com/p/ai-risk-agi-
risk
…