Concepts & ideas
Apply basic Cognitive Science concepts and ideas to explain major approaches and current developments in AI.
UNIT 5 · 2 WEEKS · 8 CONTACT HOURS
Artificial Intelligence:
from thinking machines to agents & world models
A self-paced Cognitive Science journey for students from any background. No advanced mathematics or programming experience is assumed.
Week 11 h lecturer + 3 h self-learning
Week 21 h lecturer + 3 h self-learning
Assessment6 auto-marked activities × 10 points
YOUR DESTINATION
Three learning objectives
Human ↔ machine
Infer similarities and differences between human cognition and machine intelligence through simple experiments and interactive activities.
Ethics & problem solving
Demonstrate responsible application of AI by evaluating benefits, limitations, risks and ethical implications within your area of specialization.
BEFORE THE TECHNICAL PART
Why should I learn programming or technical AI?
You do not need to become a software engineer. You do need enough technical understanding to ask better questions, design better studies, recognize what AI can and cannot do, and work with technical specialists instead of treating AI as magic.
WEEK 1
How did AI learn to behave intelligently?
Foundations → learning → programming as structured problem solving.
4 h1 h with lecturer
3 h self-paced
3 h self-paced
Lecturer hour · suggested use
Kick-off · “What counts as intelligence?” mini Turing activity · course navigation · explain how scores are recorded.
WEEK 2
Where is AI going now?
Agents → world models/JEPA → multimodal/embodied AI → responsible leadership.
4 h1 h with lecturer
3 h self-paced
3 h self-paced
Lecturer hour · suggested use
Frontier AI update · debate “Does performance = understanding?” · compare discipline-specific use cases · ethical synthesis.
FROM YOUR LECTURER
My group learning materials
Sign in to see notes, readings, videos, concept maps, code examples, discussions and extra 10-point exercises published for your group.
Sign in to load materials assigned to your group.
RESEARCH SPOTLIGHT · AUGUST 2026
Can visual experience lead toward general intelligence?
Visual General Intelligence: A White Paper
Kataoka and 20 co-authors ask whether intelligence grounded in images, video and geometry could provide another pathway toward general intelligence, rather than treating language as the only centre of AI development.
Important: this is a white paper proposing research directions and questions. It does not prove that visual intelligence has achieved AGI.
Open paper ↗Connect it to Cognitive Science
- Perception: What can vision provide that text alone cannot?
- Representation: What should an intelligent system preserve from visual experience—pixels, objects, relations, events or possible actions?
- Learning: Does learning from video resemble anything about how humans learn from continuous experience?
- Embodiment: Is seeing enough, or does intelligence require acting in and receiving feedback from the world?
- Measurement: What benchmark would convince you that a system has broad visual intelligence rather than narrow recognition skill?
AI-audit prompt: Ask an AI to summarise the paper, then verify whether it correctly distinguishes the authors' proposed research agenda from established evidence. Mark each sentence: paper-supported, interpretation, or unsupported extension.
UPDATED 7 SEPTEMBER 2026
Frontier AI radar
Current examples are used as case studies—not as proof that any company has “solved intelligence.”
MY LEARNING RECORD
Scores & progress
Sign in as a student to save results across devices. Each activity is worth 10 points.
UPLOAD TO eLEAP
Visual self-evaluation
Compare your confidence before and after the SIM, then download the visual as a PNG for your eLEAP submission.
This is a metacognitive self-evaluation, not an auto-marked quiz. Your six scored activities remain A1–A6.
READ FURTHER
Core readings & source trail
Use these to verify claims. Current product announcements are separated from independent/research sources.