Social Sciences in China (Chinese Edition)

No. 8, 2024

 

Artificial Intelligence and the Challenges to Human Thought

(Abstract)

 

Zhao Tingyang

 

Contemporary artificial intelligence (AI) employs Bayesian methodology, thereby reintroducing the problem of empiricism. While transcendental theory was unsuccessful in resolving Hume’s problem of causation, Bayesian Analysis offers an empirical solution. At the heart of Hume’s problem lies the uncertainty of the “future.” AI simplifies and reduces the abstract future into an operable “prediction of the next identifier,” ensuring the validity and credibility of predictions using Bayesian Analysis without relying on a priori methods. This compels us to reflect on human thought itself and reconsider the classic dilemmas of empiricism and transcendentalism. It is anticipated that developing a verb logic based on empiricism could potentially explain more cognitive secrets and add an analytical and predictive dimension to AI.