Artificial General Intelligence (AGI) Show with Soroush Pour Ep 1 - When will AGI arrive? - Logan Riggs Smith (AGI alignment researcher)
Nov 26, 2022
AI Snips
Chapters
Transcript
Episode notes
AGI Alignment Research
- AGI alignment research focuses on ensuring AGI safety.
- Logan Riggs Smith got into it after realizing AGI's potential catastrophic impact.
AGI Possibility
- Humans are likely Turing complete and inefficient processors compared to silicon.
- Significant advancements suggest AGI could arrive within this decade.
LLM Limitations
- Large language models are trained on text only once and lack agency in choosing input data.
- Humans have multi-modal data, agency, and can choose their learning focus, unlike LLMs.