The Optical Lattice clock will lose one second every 15 billion years.

How do we know?

In: 0

In this case the clocks isn’t displaying a time. It’s counting “tics”. Using math, they can determine how often these tics occur and how many should occur in one second. They can also measure the extremely tiny variations in how often these tics occur. They aren’t EXACTLY on time, but the variations are extremely small. So small that the total number of tics it would count in 15 billion years is only going to be off by the number they calculated occur in one second.

The smaller the variation in these “tics”, the more accurate the clock is. The counted number will closer to the expected number over time. They can also start counting at a specific time and use the number they have counted to calculate what the current time is based on the number of tics that have occurred.

In this case the tics aren’t a series of mechanical gears turning a second hand, but a measurement of the vibrations of cesium atoms.

You make multiple clocks and regularly compare them with each other at high precision. You need to very carefully compensate for differences caused by relativistic effects. Even small changes in altitude or the local strength of gravity will be noticeable with clocks of this accuracy.

Today, the most precise clocks are based on a natural atomic resonance of the cesium atom—the atomic equivalent of a pendulum.

[src](https://www.laserfocusworld.com/test-measurement/research/article/16556429/alloptical-atomic-clock-may-be-worlds-most-precise)