MKLab
Can AI Follow In Einstein's Footsteps? - Printable Version

+- MKLab (https://mklab.gr)
+-- Forum: [INDEX] (https://mklab.gr/forumdisplay.php?fid=1)
+--- Forum: ARTFICIAL INTELLIGENCE (AI) (https://mklab.gr/forumdisplay.php?fid=5)
+---- Forum: ARTICLES (https://mklab.gr/forumdisplay.php?fid=33)
+---- Thread: Can AI Follow In Einstein's Footsteps? (/showthread.php?tid=1541)



Can AI Follow In Einstein's Footsteps? - mklabgr - 08-10-2026

Can AI Follow In Einstein's Footsteps?

Summary

The paper “Can AI Follow In Einstein’s Footsteps?” examines whether AI can move beyond being an extremely powerful prediction tool to becoming a genuine engine of fundamental physics discovery. The authors argue that the historical development of physics progressed from pattern recognition and empirical laws toward increasingly general, principle-based theories such as relativity and the Standard Model, whereas AI development has largely followed the reverse path—from equation discovery and symbolic regression toward highly accurate but often opaque predictive systems such as AlphaFold and GraphCast. 

The central concern is that AI may become excellent at predicting phenomena within existing frameworks while remaining unable to formulate the right questions, new principles, or radically different theoretical frameworks needed for breakthroughs such as quantum gravity. The authors suggest that future AI systems should therefore develop capabilities involving theoretical creativity, mathematical abstraction, identifying fundamental principles, and designing experiments capable of falsifying new theories, potentially allowing AI to progress from predicting known physics to proposing the next paradigm-changing theory. 

ARTICLE