Mathematicians will probably become obsolete before anyone else [pdf]

The thread spent less time evaluating whether mathematicians are actually near obsolescence than correcting the framing: the document is a handwritten pre-LLM letter by Ted Kaczynski, not a contemporary AI forecast. That discovery made many readers treat its apparent prescience as either a coincidence or another example of vague technological pessimism being retrofitted to modern generative AI. Several commenters rejected the leap from automated formal proofs to replacing mathematicians, arguing that valuable mathematics requires humans who can judge significance, understand implications, and connect results to wider knowledge rather than merely verify symbolic output. There was some recognition that AI is already automating slices of expert work and may shrink paid roles, especially where tasks are standardized or amenable to computation. But the stronger current was skeptical: impressive output often dazzles non-experts more than practitioners, and machine-generated mathematics without human comprehension is not a substitute for human mathematical culture or insight. A side debate over whether programming belongs to mathematics reinforced the same distinction between using computational tools and doing foundational mathematical work. Readers also raised a broader risk that AI-transformed knowledge jobs could become centrally controlled through opaque, closed systems, leaving fewer people able to inspect or challenge the results.