Cal Day Lecture: Can a Machine be Conscious? Towards a Conscious AI
summary
Defines a Conscious Turing Machine whose tiny short-term memory broadcast sends a single small “stage” to ~10^11 unconscious processors; each processor proposes triples (address, content, weight) and competing weights slowly return one chunk to the stage (George Miller’s 7±2 chunks). Shocking pain is an interrupt that forces every processor’s attention; pain asymbolia shows simulation vs. suffering. Titan-class machines (~1.7×10^15 transistors vs. ~1.3×10^15 synapses) might be large enough; no CTM is built yet.
AI-written summary · about
chapters (13)
- 0:00Defining consciousness
- 3:10The hard problem
- 7:37Conscious Turing machine
- 10:24The theater of the mind
- 13:38Working memory and input
- 17:37Architecture of processing
- 22:45Dynamic system modeling
- 28:23Chunks and attention
- 34:32Core consciousness factors
- 36:56Pathologies and limitations
- 41:24Engineering emotions
- 45:53Engineering pain
- 52:13Q&A and closing thoughts
related talks
45mThe Conscious Turing Machine (CTM), a formally defined Theoretical model of Consciousness
47mWhat is Consciousness and Can Machines have it? Part 1
01h 14mAI Consciousness is Inevitable: A Theoretical Computer Science Perspective
16mWhat is Consciousness and Can Machines have it? Part 2
43mThe Role of TCS in Understanding (AI) Consciousness
21mSentient AI – Can We Create Artificial Consciousness?
38mA rocky road to general AI
09mSmale’s 18th Problem and Intelligence Science 2