911±¬ÁÏÍø

Aims:

The aims of the "Open-Endedness and General Intelligence" module are to:

• Provide students with a comprehensive understanding of advanced AI research, focusing on the development of open-ended and generally-capable AI systems.
• Equip students with a strong foundation in open-ended learning principles, techniques, and methodologies.
• Encourage a culture of innovation and creativity, empowering students to explore novel ideas and approaches in AI research and development.

Intended learning outcomes:

On successful completion of the module, a student will be able to:

1. Demonstrate a comprehensive understanding of the principles, techniques, and methodologies underpinning open-endedness and general intelligence in Artificial Intelligence.
2. Analyse and criticize current Artificial Intelligence research and breakthroughs, reflecting on their implications for the development of more agentic, generally-capable Artificial Intelligence systems.
3. Synthesize practical Artificial Intelligence solutions for various domains, such as robotics and language processing.
4. Engage in innovative and creative problem-solving, utilizing novel ideas and approaches in Artificial Intelligence research and development.

Indicative content:

The following are indicative of the topics the module will typically cover:

• Foundation models, large language models, world models.
• Techniques for promoting exploration and intrinsic motivation in AI agents.
• Optimization approaches such as novelty search, quality diversity algorithms and evolutionary computation.
• Automated curriculum learning.
• Self-referential learning and self-improvement.Ìý

Requisite conditions:

To be eligible to select this module as optional or elective, a student must: (1) be registered on a programme and year of study for which it is formally available; and (2) have attended Supervised Learning (COMP0078), and either Applied Deep Learning (COMP0197) or Bayesian Deep Learning (COMP0171) in Term 1.