41 ms·

This sort of change is a great idea on a fundamental level: define our units in terms of fundamental constants or processes that are absolutely reproducible. Bu

by Steuard 13y ago

This sort of change is a great idea on a fundamental level: define our units in terms of fundamental constants or processes that are absolutely reproducible. But as the article explains, any such definition has to be at least as precise in practice as the less fundamental definitions that they replace. Hence the challenge (mentioned in the article) of defining the kilogram in a sensible way. The obvious method would be to say "1 kg = the mass of [###] of carbon atoms" or something like that. But counting 10^23 individual atoms is impossible for all practical purposes: it's still less precise than just weighing that reference lump of metal (despite the flaws of that approach). The remarkable thing described in this article is that scientists are evidently at the verge of being able to count one electron at a time fast enough for use in calibrating actual, practical current measurement devices. That's pretty cool.

arethuza 13y ago
"Feynman, I know why all electrons have the same charge and the same mass" "Why?" "Because, they are all the same electron!" http://en.wikipedia.org/wiki/One-electron_universe http://en.wikipedia.org/wiki/One-electron_universe
VLM 13y ago
"The obvious method" Using a cesium time standard and an atomic force microscope figure out the half life of individual atoms of 14C (or something more fun/shorter) then define a kilo of 14C as a pile of 14C generating X decay events per whatever interval calibrated by our cesium time standard. (edited to add the point is you can watch individual atoms with the AFM and see which ones disintegrated in the past since the last scan, and its no great trick for a century or so to detect individual decays using a geiger counter. Also 14C is no fun for this task. Try a shorter halflife)
- bcbrown 13y ago
  The half-life is statistical, not mechanistic, so there would be an inherent variability in that definition. The larger the radioactive mass and the longer the time period, the less variability, but it's still variable.
  - jjoonathan 13y ago
    The fact that it's variable isn't a valid criticism on its own: there is uncertainty in every standard we use, and the objective is to minimize that uncertainty. It's quite possible that the practically achievable error bars from this method would be thinner than the practical error bars of another method. Besides, the fact that you can arbitrarily increase the precision of this method without hitting any hard physical limitations is a solid selling point.
    - bcbrown 13y ago
      You are right that there's always going to be uncertainty. Perhaps it turns out that in practice, the uncertainty with this method is less than other approaches. The point I was trying to make was about the different between incidental uncertainty due to imprecise equipment, and more fundamental uncertainty due to physical laws. For example, other intrinsic uncertainties are the Planck resolution for time and space, and the Heisenberg Uncertainty Principle for certain combinations of information. Similar to those, radioactivity has a fundamental limit. We cannot predict when an atom will emit radiation, all we can do is make a statistical prediction.

by Steuard 13y ago

arethuza 13y ago

"Feynman, I know why all electrons have the same charge and the same mass" "Why?" "Because, they are all the same electron!" http://en.wikipedia.org/wiki/One-electron_universe http://en.wikipedia.org/wiki/One-electron_universe

VLM 13y ago

"The obvious method" Using a cesium time standard and an atomic force microscope figure out the half life of individual atoms of 14C (or something more fun/shorter) then define a kilo of 14C as a pile of 14C generating X decay events per whatever interval calibrated by our cesium time standard. (edited to add the point is you can watch individual atoms with the AFM and see which ones disintegrated in the past since the last scan, and its no great trick for a century or so to detect individual decays using a geiger counter. Also 14C is no fun for this task. Try a shorter halflife)

bcbrown 13y ago
The half-life is statistical, not mechanistic, so there would be an inherent variability in that definition. The larger the radioactive mass and the longer the time period, the less variability, but it's still variable.
- jjoonathan 13y ago
  The fact that it's variable isn't a valid criticism on its own: there is uncertainty in every standard we use, and the objective is to minimize that uncertainty. It's quite possible that the practically achievable error bars from this method would be thinner than the practical error bars of another method. Besides, the fact that you can arbitrarily increase the precision of this method without hitting any hard physical limitations is a solid selling point.
  - bcbrown 13y ago
    You are right that there's always going to be uncertainty. Perhaps it turns out that in practice, the uncertainty with this method is less than other approaches. The point I was trying to make was about the different between incidental uncertainty due to imprecise equipment, and more fundamental uncertainty due to physical laws. For example, other intrinsic uncertainties are the Planck resolution for time and space, and the Heisenberg Uncertainty Principle for certain combinations of information. Similar to those, radioactivity has a fundamental limit. We cannot predict when an atom will emit radiation, all we can do is make a statistical prediction.