I begin with developments in artificial intelligence and move toward the philosophical core.<\/p>\n
Between 2020\u20132023, AI systems based on symbolic regression emerged.<\/p>\n
One notable example was developed by researchers at ETH Zurich and MIT.<\/p>\n
The idea: feed AI only experimental data \u2014 and it attempts to extract the mathematical equation describing it.<\/p>\n
Remarkably, it rediscovered Newton\u2019s laws, re-derived gravity equations, and extracted relations in complex astronomical systems.<\/p>\n
It did not \u201cunderstand\u201d physics.
It searched for the simplest mathematical structure explaining the data.<\/p>\n
In 2023, researchers published in Nature<\/em> about AI discovering:<\/p>\n Hidden conservation laws<\/p>\n<\/li>\n Equations describing double pendulum motion<\/p>\n<\/li>\n Chaotic systems difficult to derive manually<\/p>\n<\/li>\n<\/ul>\n The AI was not given the law \u2014 it inferred it from data.<\/p>\n Did it prove them? Named after Richard Feynman, the system aimed to discover physical laws automatically.<\/p>\n It successfully rediscovered more than 100 physics equations from data alone \u2014 a genuine scientific achievement.<\/p>\n So can AI discover entirely new laws? The philosophical point: AI does not create from nothing; it extracts latent structure from data.<\/p>\n This raises a deeper question: AI discovering new equations is not a trivial technical event. Now a third entity enters: a computational system without feeling or contemplation \u2014 yet reaching correct conclusions.<\/p>\n Does knowledge require consciousness? Is mind a biological property \u2014 or a universal structure?<\/p>\n Perhaps mind is not matter, but information organization.<\/p>\n AI may solve \u2014 but it does not witness.<\/p>\n Is witnessing required for truth?<\/p>\n","protected":false},"excerpt":{"rendered":" I begin with developments in artificial intelligence and move toward the philosophical core. 1. Symbolic Regression Between 2020\u20132023, AI systems based on symbolic regression emerged. One notable example was developed by researchers at ETH Zurich and MIT. The idea: feed AI only experimental data \u2014 and it attempts to extract the mathematical equation describing it. …<\/p>\n","protected":false},"author":1,"featured_media":12724,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[280,269,103,16,197],"tags":[],"class_list":["post-13625","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-2026-collective-activities-articles","category-al-hurriya-website","category-all-articles","category-articles-by-dr-badrawi","category-translated-articles"],"yoast_head":"\n\n
AI proposes equations, tests them numerically, verifies alignment with data.
Formal mathematical proof is still performed by humans.<\/p>\n3. AI Feynman (2019)<\/h3>\n
Yes \u2014 early evidence suggests so.<\/p>\n
Are laws inherent in nature waiting to be discovered?
Or are they human mathematical formulations making nature comprehensible?<\/p>\n
Historically, discovery was a purely human act \u2014 a conscious mind confronting mystery.<\/p>\n
Or is correctness enough?<\/p>\n