What this course actually covers
Most AI courses stop at deep learning benchmarks and task-specific models. This one starts where those leave off. We examine what separates narrow AI systems from general-purpose reasoning — not as a philosophical exercise, but as a technical problem with measurable dimensions.
You will work through core frameworks: integrated information theory, global workspace theory, and symbol-connectionist hybrid architectures. Each week pairs a theoretical model with a hands-on case study, so the abstractions stay grounded.
Who benefits most
This course suits ML engineers who want to move beyond fine-tuning pipelines, researchers transitioning from cognitive science into AI, and software architects evaluating AGI-adjacent systems. If you have never trained a neural network, start with our prerequisite module first.
What you leave with
By the end, you will be able to read current AGI research papers critically, identify the architectural assumptions behind different AGI proposals, and articulate where today's large language models fall short of general reasoning. That is a concrete, testable skill set — not a vague familiarity with the topic.
AGI Foundations: From Narrow AI to General Intelligence
What this course actually covers
Most AI courses stop at deep learning benchmarks and task-specific models. This one starts where those leave off. We examine what separates narrow AI systems from general-purpose reasoning — not as a philosophical exercise, but as a technical problem with measurable dimensions.
You will work through core frameworks: integrated information theory, global workspace theory, and symbol-connectionist hybrid architectures. Each week pairs a theoretical model with a hands-on case study, so the abstractions stay grounded.
Who benefits most
This course suits ML engineers who want to move beyond fine-tuning pipelines, researchers transitioning from cognitive science into AI, and software architects evaluating AGI-adjacent systems. If you have never trained a neural network, start with our prerequisite module first.
What you leave with
By the end, you will be able to read current AGI research papers critically, identify the architectural assumptions behind different AGI proposals, and articulate where today's large language models fall short of general reasoning. That is a concrete, testable skill set — not a vague familiarity with the topic.
Program structure
What gets covered and in what order — no filler, no repetition.
Course Structure
- Week 1 — Defining AGI: benchmarks, Turing-style tests, and their limitations
- Week 2 — Cognitive architectures: ACT-R, SOAR, and OpenCog compared
- Week 3 — Memory and reasoning: episodic vs semantic systems in biological and artificial agents
- Week 4 — Transfer learning and meta-learning as stepping stones toward generality
- Week 5 — Self-supervised world models and predictive coding
- Week 6 — Safety constraints and alignment considerations in general systems
- Week 7 — Research paper deep-dive and peer critique session
- Week 8 — Final project: architectural proposal with documented trade-offs
Each session includes a recorded lecture, a reading list of 2 to 4 papers, and a structured discussion thread moderated by the instructor.
Course methodology note
Ready to start?
Seats fill up quickly — once the cohort closes, the next opening is months away.