Hi! I'm Zory Zhang,
Here is my resumé:
Hope you like here >_<
You’re very welcome to reach out and talk about anything!
Me @Hong Kong West Kowloon Cultural District, 2022 Jan
I’m interested in human-like learning and reasoning. I attmpt to computationally characterize and model human reasoning by asking what does it take to build (reverse engineering) a human-like (artificial) cognitive system: A system that can learn sample-efficiently and continually without catastrophic forgetting and flexibly generalize during performance.
I believe that reasoning is not only a performance task but also help human-like learning, i.e. reasoning to learn (Reason2Learn). Given prior belief in nearby domain, humans are able to learn from only a few examples in a novel domain and infer knowledge about the novel domain. This mission bridges artificial intelligence and cognitive science, producing a flexible, explainable and persuasive system as a byproduct. My toolkit includes hierarchical concept formation (cobweb), schema induction and explanation-based learning for human-like learning, analogical reasoning and “neuro-symbolic” architecture for human-like inference, as well as employing formal mathematics theorem proving as a playground to study problem solving and reasoning.
(Note that the postulate that human intelligence is a product of evolution given biological constraints suggests that an alignment between computational system and human is precious but not necessary. My way of avoiding some ethical problems of AI is to build zombie AI, i.e. try not to let AI have consciousness, but stay merely a sophisticated system. Another viewpoint is that dangerous AI may act as an external threat that leads to the unity of humanity.)
PREPRINTS
Writing is the process through which ideas are produced and refined.
In this space, I will post blogs that contain mostly my original articulation, therefore prioritized over other pages on my website, which are mainly notes.