Pedro Domingos, University of Washington professor emeritus, frames computer science around tractable problems that evaluate functions and intractable ones that invert them, positioning AI in the latter category

— Pedro A. Ortega reframed the split as deduction versus induction.

Original post

OPPD#654Pedro Domingos|@PMDDOMINGOS

There are only two real complexity classes in computer science: tractable (computing a function's output from its inputs) and intractable (computing a function's inputs from its output). AI deals with the latter. The rest is details.

6:29 PM · May 19, 2026 View on X

QUOTE POSTPA#1258Pedro A. Ortega|@ADAPTIVEAGENTS

The two classes are effective deduction and the effective induction, more commonly known as computation and artificial intelligence, respectively

PDPedro Domingos@pmddomingos

There are only two real complexity classes in computer science: tractable (computing a function's output from its inputs) and intractable (computing a function's inputs from its output). AI deals with the latter. The rest is details.

1:29 AM · May 20, 2026 · 5.1K Views

9:10 AM · May 20, 2026 · 1K Views

REPLYPA#1258Pedro A. Ortega|@ADAPTIVEAGENTS

AI can be reduced to computation if you supply nondeterministic side information

PAPedro A. Ortega@AdaptiveAgents

The two classes are effective deduction and the effective induction, more commonly known as computation and artificial intelligence, respectively

9:10 AM · May 20, 2026 · 1K Views

9:14 AM · May 20, 2026 · 137 Views