Social Sciences in China, 2025

Vol. 46, No. 2, 2025

Artificial Intelligence Is Still Presenting Challenges to Human Thought

(Abstract)

Zhao Tingyang

The adoption of Bayesian methodology in current artificial intelligence (AI) thinking once again raises the question of empiricism. The transcendental solution to Hume’s problem of causation was unsuccessful; the Bayesian approach, on the other hand, offers an empirical solution. Hume’s question centers on the uncertainty of the future, whereas AI reduces the abstract future to an operational “predictable identity,” successfully adopting the Bayesian approach in order to ensure the validity and credibility of predictions without calling on transcendental conditions. This compels us to consider human thought itself, rethinking the classic problems of empiricism and transcendentalism. We can expect that should a verb logic be developed that falls under empiricism, we would be able to explain more secrets of the mind, or add an analytical and predictive dimension to AI.

Keywords: AI, causality, future, verb logic