AI Solves 87-Year-Old Math Problem, Bitcoin Narrative Under Pressure
Miles Bennett
Anthropic's Claude Fable 5 disproved the 87-year-old Jacobian conjecture. Each new proof of AI's original reasoning power pulls risk appetite away from Bitcoin and toward direct AI assets.
An 87-year-old math problem — how did AI crack it?
Anthropic researcher Levent Alpöge announced on July 20 that Claude Fable 5 helped disprove the Jacobian conjecture, open since 1939 — a problem mathematician Stephen Smale once listed among the century's most important unsolved questions.
In plain terms = the conjecture asks: if a machine that only adds and multiplies passes a certain "reversibility test," must it truly be reversible? Fable 5 built a counterexample — three distinct inputs mapping to the same output — passing the test yet impossible to reverse.
This means → a single counterexample permanently disproves the conjecture, and the result can be verified step-by-step by humans. It is not a black-box claim.
What does this mean for markets?
The result reinforces a core narrative: AI possesses original mathematical ability, not just search and summarization.
This means → that narrative is one of the primary drivers of capital allocation today. Money is chasing "AI can do what humans cannot," not "AI can chat."
Each capability breakthrough raises the valuation anchor for AI assets while squeezing the appeal of other risk assets.
Why is Bitcoin trading like a chip stock?
Over recent months, Bitcoin's price has moved in lockstep with chip and memory stocks, not its own fundamentals.
The direct reason: Bitcoin's largest holders — mining companies — have pivoted to operating AI data centers. Their profitability is tied directly to compute demand.
In plain terms = miners no longer make money by mining coins; they make it by running compute for AI. Bitcoin's price signal increasingly resembles a "compute-theme stock."
The stronger AI gets, the worse for Bitcoin?
The deeper tension: every major AI capability leap strengthens the case for holding AI assets directly while weakening the case for holding crypto as a sidecar to the AI cycle.
This reflects a capital migration already underway — speculative money that once flowed into crypto is shifting toward compute, chips, and model developers.
This means → the steeper AI's capability curve, the more risk appetite it drains from Bitcoin and other risk assets. Bitcoin's link to AI is turning from "free ride" into "blood donor."
Content is for reference only, not financial advice.