There’s still no regulation on what you can do with machine learning (At least until your deepfakes constitute unlawful use of someone’s likeness, or you use your classifier to make hiring decisions or something, or something like that.)
Machine learning is a technique of mathematics. What do you even mean there’s no regulation of this subset of mathematics? If the law were to say a triangle can’t have three sides would it make it so?
"Since 1967, the second has been defined as exactly "the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium-133 atom" (at a temperature of 0 K)."
Therefore, the meter may be defined as "the distance light travels in a vacuum in one of those cesium-133-related periods, times 9,192,631,770, divided by 299,792,458".
Right, but the constant c didn't have that exact value until 2019. Prior to 2019, c had an infinite number of digits. Thus, the numerical representation of constants with units can change as the definitions of the units themselves change.
I don't follow your thought process here. The speed of light has units, that's true. Being a speed, it has units of distance over time. But the question "how can you define the meter in terms of the speed of light if the speed of light depends on the definition of the meter?" is totally incoherent; the premise is wrong. The speed of light does not depend on the definition of the meter, making it very easy to define the meter in terms of the speed of light.
To take a trivial example, the speed of light is 300,000,000 meters per second and 300,000 kilometers per second. If we relabeled kilometers "flags", then the speed of light would also be 300,000 flags per second. But while 300,000,000 and 300,000 are different numbers, 300,000,000 meters and 300,000 flags are the same distance. You can't change the speed of light by defining a new unit of length. You can only change the coefficient of that unit that gives the speed of light. But coefficients are dimensionless and therefore aren't speeds. When you multiply the coefficient by the unit, you get the same constant speed as always.
I can't tell why you think it's "problematic" for a physical constant to have units, but I feel safe in saying that whatever you have in mind, it's wrong.
I agree, I messed that up. What I meant to say was: "how can the speed of light have an unchanging numerical value if the definition of the meter is not an unchanging definition?" Prior to 2019 the meter was not defined in terms of speed of light over time.
My point is that, unlike dimensionless constants like pi, constants with units can have arbitrary numerical values depending on the definitions of the units. And if you make a law that changes the definition of a unit, you therefore change the numerical representation of the corresponding constants.
In other words, if you defined a meter to be "the distance light travels in 1 second", then the constant c would now be 1 m/s.
> I can't tell why you think it's "problematic" for a physical constant to have units
Because the definitions of units can change[1]. The definition of the meter changed earlier this year, and thus the speed of light constant, c also changed. Before May 9, 2019, c had an infinite number of digits using SI units. Now it only has 9 digits. Planck's constant, along with a bunch of others changed as well during the units redefinition[1].
[1] https://en.wikipedia.org/wiki/2019_redefinition_of_the_SI_ba...
Again, I'm not arguing that the speed of light can change, but that those constants are only unchangeable in their symbolic forms. Laws can and do change their numeric representations by proxy of defining units.
As I and you have already stated, the speed of light cannot have any purely numerical value, because it is a dimensional quantity.
...and? What's the problem supposed to be? I've never seen anyone be confused over the idea that the speed of light is 300,000,000 when measured in meters per second, but 186,000 when measured in miles per second.
Read the whole thread, start to finish and the context should make it clear.
> I've never seen anyone be confused over the idea that the speed of light is 300,000,000 when measured in meters per second, but 186,000 when measured in miles per second.
Right, because unit conversion is not confusing. But what can be confusing is unit re-definition, which doesn't happen very often and can seem counter-intuitive at first ("how can a government seemingly change the value of a natural constant?!").
Many well known examples in the encryption field.
While we're at it, whenever anyone talks about machine learning they're not talking about a technique of mathematics, they're talking about a technique in software development (albeit that has some mathematical roots.) It's not popular because we're all passionate about discovering mathematical truths, it's popular because we can use it to make computers do cool things. :)