Positive Lower Density for Hofstadter's $ab-1$ Problem

The Infinite Game of Number Building

Imagine a vast, endless playground where numbers are the toys. In the world of mathematics, specifically a branch called number theory, researchers love to play games with rules that generate new numbers from old ones. One of the most famous types of games involves "recurrence" or "renewal." Think of it like a game of musical chairs, but instead of people, it's numbers, and instead of a chair, it's a specific spot on a number line. The big question mathematicians have been asking for decades is: if you keep playing this game forever, do the numbers you create spread out evenly across the playground, or do they clump together in one corner, leaving huge empty spaces?

This specific paper tackles a puzzle that started with a simple rule: start with the numbers 2 and 3. Then, take any two different numbers you already have, multiply them, and subtract 1. If the result is a whole number, add it to your collection. Repeat this forever. The question, posed by the legendary mathematician Paul Erdős (who heard it from the author of the famous "Hofstadter's Figure-Figure" sequences), is whether this collection of numbers is "thick" enough. Does it have a "positive lower density"? In plain English, does the set of numbers you generate eventually fill up a significant, non-zero percentage of the number line, no matter how far out you go? For a long time, no one knew if the answer was yes or no.

The Solution: A Traffic System for Numbers

In this paper, Samuel Korsky proves that the answer is yes. The set of numbers generated by this rule does indeed have a positive lower density. This means that as you look at larger and larger ranges of numbers, you will always find a guaranteed, non-zero chunk of them belonging to this special set. It's not just a few scattered numbers; they are abundant.

To understand how the author solved this, imagine the set of numbers as a city, and the rule "multiply and subtract 1" as a set of one-way streets. The author's goal was to show that there are so many different ways to drive through this city that you can't avoid hitting a lot of destinations. However, there's a catch: the rule says you can only multiply distinct numbers. If you try to multiply a number by itself, the rule breaks. This is like a traffic law that says you can't drive on a road if you've already been on that exact same road segment in the same trip.

The author's strategy is to build a "traffic control system" using a map divided into 20 specific zones (intervals). He assigns different "multipliers" (like 2, 3, 5, 9, 14) to these zones. When a number lands in a zone, the system tells it which multiplier to use next. The genius of the proof lies in how these multipliers are chosen. The author sets up four different "traffic patterns" (assignments). By switching between these patterns based on the current state of the system, he ensures that the numbers don't get stuck or run into the "distinctness" rule violation.

Think of it like a game of "Follow the Leader" where the leader is trying to keep a perfect balance. The author tracks the "ingredients" of the numbers (specifically the powers of the prime numbers 2, 3, 5, and 7). He wants the recipe to stay balanced so that the numbers grow in a very specific, predictable way. He uses a feedback loop: if the recipe gets too heavy on the number 2, the system switches to a pattern that adds more 3s or 5s to balance it out. This keeps the "slope" of the growth (how fast the numbers get bigger) locked onto a specific target.

The paper shows that by carefully managing these switches, the system creates a massive number of unique paths that all end up at the same "slope." Because the paths are unique and the system is designed to return to its starting point over and over again (a concept called "positive recurrence"), the math proves that there are infinitely many distinct numbers generated.

Crucially, the author proves that these paths are distinct even though the underlying math allows for some overlaps (the system isn't "free" in the strict mathematical sense). He does this by showing that if you trace the paths backward on his 20-zone map, they never cross each other until they reach the very end. This guarantees that every path produces a unique final number.

The final punchline is a counting argument. The author calculates that for every "step" in this process, the number of valid paths grows at a rate that matches the growth of the numbers themselves. He proves that for a specific large number $x$, the amount of his special set found within $1$ to $x$ is at least $c \times x$, where $c$ is a constant greater than zero. In other words, no matter how far you count, you will always find a steady stream of these numbers.

The paper doesn't just suggest this is likely; it provides a rigorous, step-by-step mathematical proof. It uses a combination of probability (to show the system keeps returning to its start), geometry (to map out the intervals), and number theory (to count the prime factors). The result is a definitive answer to a decades-old question: the set is not sparse; it is dense, filling the number line with a reliable, positive presence.