PERSON

Alfred Tarski

The logician who gave the twentieth century its only rigorous, paradox-free account of what “true” means—and proved, in the same stroke, that no sufficiently expressive system can define its own truth from within, a limit that maps with uncomfortable precision onto every serious question about whether AI systems can certify their own outputs.

We have built machines that produce sentences faster and more fluently than any institution in history, and almost no one stopped to ask the question Alfred Tarski spent his life answering: what does it mean for a sentence to be true? Born Alfred Teitelbaum in Warsaw in 1901, trained under Stanislaw Leśniewski and elevated to the Lwów–Warsaw School of mathematical logic before he was thirty, Tarski sailed for the United States in August 1939 to attend a conference and could not go home; most of his family was murdered in the years that followed. He rebuilt himself at Berkeley and made it the world’s foremost center of mathematical logic, training a generation before his death there in 1983. In 1933 he published the result that organized his career: a mathematically rigorous definition of truth for a formal language, along with a proof that

Keep reading with YOU ON AI