In this talk, @BuzzRobot guests Igor Grossmann and Sam Johnson from the University of Waterloo talk about improving AI reasoning from a cognitive science perspective: how implementing wisdom and metacognition in AI systems will help them develop further abstract thinking and reasoning capabilities, and make them more robust and prepared to operate in unpredictable and new environments.
Timestamps:
0:00 Story behind this research work
1:18 Introduction to the research
3:58 Current shortcomings of AI systems
5:31 Examples of human wisdom
10:05 Why metacognition is central to wisdom
17:51 Towards wise AI systems
18:59 Benefits of wise AI systems
25:23 Embracing metacognition in AI training
27:11 Challenges of building wise AI systems
28:12 Suggestions for building metacognition into AI systems
#artificialintelligence #artificialgeneralintelligence #machinelearning #deeplearning #googledeepmind #openai #chatgpt #largelanguagemodels #largelanguagemodel #llms #aiagents #agi #cognitivescience #phylosophy #technology #science #techtalk #airesearch #software #metacognition #wisdom
Social Links:
Newsletter: https://buzzrobot.substack.com/
X: https://x.com/sopharicks
Slack: https://join.slack.com/t/buzzrobot/shared_invite/zt-2s067rv7n-guPIMGe62rbp9ncxdnOUfQ