Does AGI require core knowledge? An AI cognitive benchmark

AI models, especially LLMs, are generally focused on achieving performance in high-level cognitive skills. But what about fundamental cognitive abilities that humans already aquire early in life? The Core Knowledge Benchmark investigates these fundamental abilities in AI.

For humans, learning rudimentary concepts (e.g., object hierarchy) represents a building block towards developing higher cognitive skills. What are the consequences of having AI that lacks these rudimentary skills while trying to perform more complex cognitive tasks (e.g., language)?

The Core Knowledge Benchmark:

1. Evaluates AI performance on a number of rudimentary cognitive skills.
2. Explores the current deficit in AI core performance.
3. Investigates the relation between core abilities and high-level abilities.

We’ll walk through the main cognitive concepts introduced in this paper and discuss the implications of their findings on achieving AGI with or without core human abilities.

Who should attend?

This event is for anyone seriously interested in the frontiers of AI and AGI, whether you’re a researcher, engineer, student, or thoughtful enthusiast. Expert level knowledge is not required. Curiosity and minimal preparation are enough.

Plan for the evening

We meet in a large space rented by one of our co-organizer. It's ideal for deep conversations.

Here is the rough plan for the evening:

• 🍕🍻 18:30–19:00 – Arrival, pizza & drinks, casual networking
• 🤝 19:00–19:30 – Welcome & intro round
• 🗫 19:30–21:30 – Group discussion

Looking forward to a great and stimulating discussion!

Cheers!
– Alexandra, Dibya, Nico & Somayeh