AI Needs Metacognition to Master Both Fast and Slow Thinking
Thinking Fast and Slow in AI: The Role of Metacognition
Despite rapid advances, AI remains narrow, lacking capabilities that come naturally to human intelligence. This paper proposes a multi-agent architecture inspired by Daniel Kahneman's theory of thinking fast and slow: incoming problems are solved either by system 1 agents that react using past experience, or by system 2 agents deliberately activated when reasoning and search for optimal solutions are required. Both agent types draw on a world model with domain knowledge and a self model tracking past actions and solver skills.
State-of-the-art AI still lacks many capabilities that would naturally be included in a notion of (human) intelligence.
- red75prime
It reminds me of "At the time we drew boxes labeled 'perception', 'cognition' with arrows between them." An imprecise quote that I can't place.
I guess my box labelled 'subconsciousness' is trying to say that low-level mechanisms that give rise to the observed cognitive phenomena might have nothing to do with neat boxes.
- creativeSlumber
How relevant is this fast/slow thinking thing with regards to current frontier models?
I know a large organization who's built their AI framework completely around this concept, and I feel that it's not really meaningful concept with the capabilities of current models.
- crorella
It looks like a lot like how data bases query optimizers work, with the exception that in the paper there is also a learning/memory component that conditions the evaluation of the answer provided by the first model.
- zfoong
At least this is written before ChatGPT.
- vist_orn
Trying to get LLMs to 'think about their thinking' is my daily struggle. This paper nails why it's so critical.