Hilbert's Tenth Problem, posed in 1900, asks whether there exists a general algorithm that can determine if any given polynomial equation has whole-number solutions. For integers, this was famously shown to be impossible in 1970. The natural follow-up question is whether the same holds for rational numbers, meaning fractions. This turned out to be much harder, and only very recently was it resolved, with a proof showing that no such algorithm can exist for rationals either.
The new result in this paper sharpens that conclusion in a specific technical way. The recent breakthrough established that no algorithm can decide whether polynomial equations over the rationals have solutions, but the paper asks: how many variables do you actually need for this impossibility to kick in? The authors prove that 86 variables are sufficient. In other words, just within the world of polynomials in 86 unknowns, the problem is already algorithmically unsolvable.
Why does the number of variables matter? Fewer variables generally means a more constrained and potentially more tractable problem, so pinning down a concrete small number is a meaningful refinement. It shows the undecidability is not some exotic feature requiring arbitrarily complicated equations but is already present in a fixed, modest-dimensional setting. The paper builds on the technical machinery from the recent solution, including tools called height estimates and pole-parity constructions, and tightens the argument to achieve this explicit bound of 86 variables.