Elon Musk / Tesla / SpaceX / Twitter

[lawtig02]

Whether or not it falls under a lottery, how is this not some form of fraud?

Yeah, that was the main takeaway from yesterday's hearing IMO. The state was seeking an injunction on the lottery theory. Elon's lawyer responded by essentially saying, "this was a fraudulent scam, not a lottery." The judge apparently agreed. So we'll see what happens, and I'm not sure any potential plaintiffs could establish actual damages, but it's always tricky when your best defense is that you're a just an ordinary fraudster, not a criminal.

 

[superrific]

Whether or not it falls under a lottery, how is this not some form of fraud?

Well, this is why the lawyer was making that distinction. He wants to have his cake and eat it too. But most of us can appreciate that if it's not a lottery, then it's fraud, and it's only not fraud if it's not a lottery.

 

[Ddseddse]

If you open a book without looking and point at a specific passage, is that selecting text at random? There's not a uniform probability distribution. For one thing, longer paragraphs have greater chance of being selected. Also, imagine yourself doing that at home right now. Are you going to really open randomly to page 2? And you're also likely to use your finger to point at something in the middle of the page, not marginalia.

I think most people would consider this to be random.

No. Outcomes with uneven probability distribution are, by the very definition, decidedly not random. That's why it's precisely so damned hard to create a real random number generator, Almost every observable phenomena we encounter has some order built into it. Basically the best we can do is use cosmic background radiation as a proxy for true randomness.

 

[superrific]

We may be getting twisted up between legal and mathematical definitions here, but I do not see the outcome of 2D6 as random, but rather as the product of two separate random events. The quintessential randomness is in the roll of each singular die, not in any output you compute by combining several die rolls. I would however say that that output is determined "by chance".

Like I say, have no idea about how the meaning of those words would be adjudicated in a court of law by any given judge, but that is the plain meaning of those words to me.

Welcome to the legal field of statutory interpretation, in which we try to give precise meanings to imprecise words according to a set of vague principles!

There's no question that you are drawing a real distinction. A uniform probability distribution is meaningfully different than one that is not. See technical note below. But then the trick is to figure out how that distinction maps onto our language, and that's not easy because of those words at the end of your post. "To me." Yeah, we're all stuck in that epistemic quandary to some extent.

Brief anecdote: in law school, I took a class taught by John Manning (who, rather improbably, became the Dean of the law school, and then of the whole university. It's not common that Jewishness is an asset on a resume but in this case . . . ). He and I used to spar all the time. He's a textualist; that is, he's a big fan of using only words to discern meaning. And in class, he would say things like, "and this, after all, is what this word means." And I would say "but couldn't you equally argue that the word means this other thing." His response would be, "that's clever, but isn't it really this?" Of course, that raises the question of determining what's clever and what's true. As many students in the class came to observe, that's a question of authority as much as one of language. Interpreting words can be tricky.

Anyway, back to randomness. To get around the "to me" problem, we often try to look at language how it's actually used. For instance:

1. A guy comes up to you in the grocery store parking lot. He says, "I'm starting a business. If you loan me $100, I'll pay you back $200 next year." I think most people would think it appropriate to respond, "No offense, but I don't loan money to random people." Do you? But of course, on your definition of random as uniform probability distribution, it's not random at all. You only ran into this guy because you shopped at this store and not one in a different city or even across town; because you allowed him to approach you instead of walking away briskly; because you go to the store at all, etc.

2. The victims who were killed by the bullets meant for Trump in Butler PA. We call those deaths random, right? But they weren't random at all, per your definition. For one thing, you had to have been at a Trump rally to be shot. Second, you had to in the line of fire from the specific place where a gunman could get off a shot at Trump. Most people at the event had a 0% chance of being shot.

3. When people are asked to name a random number between 1 and 100, the results are not uniform. For one thing, almost nobody guesses 2, nor multiples of 10. I watched a video about this. I think the most chosen numbers (by a considerable margin, even) are 37 and 73. There's definitely a gradient.

But if I ask you to name a number between 1 and 100, any number, I think we'd describe that as random. And if I set the humidity in a commercial greenhouse to that number, my business partner might be justifiably outraged that I just set the growing conditions to a random number pulled out of someone's butt. Saying, "well, it's more likely to be 73 than 2" would not answer the criticism.

In general, I think that people use the word randomly to denote a process in which the outcome is not fully predictable, or if an outcome (or an input) is established by reasons that are not apparent.


Technical note: describing the probability distribution of a 2d6 isn't fully explained as "two separate random events." I could weight a 12 sided die so that its probability distribution would be the same as 2d6. You could say, well that simulates the 2d6 probability, which is true in a sense; but it's also true that the 2d6 actualizes a type of probability distribution that can be actualized in other ways as well. But that's not really important and also I might not be right about this point and I don't want to argue it.

 

[Ddseddse]

Welcome to the legal field of statutory interpretation, in which we try to give precise meanings to imprecise words according to a set of vague principles!

There's no question that you are drawing a real distinction. A uniform probability distribution is meaningfully different than one that is not. See technical note below. But then the trick is to figure out how that distinction maps onto our language, and that's not easy because of those words at the end of your post. "To me." Yeah, we're all stuck in that epistemic quandary to some extent.

Brief anecdote: in law school, I took a class taught by John Manning (who, rather improbably, became the Dean of the law school, and then of the whole university. It's not common that Jewishness is an asset on a resume but in this case . . . ). He and I used to spar all the time. He's a textualist; that is, he's a big fan of using only words to discern meaning. And in class, he would say things like, "and this, after all, is what this word means." And I would say "but couldn't you equally argue that the word means this other thing." His response would be, "that's clever, but isn't it really this?" Of course, that raises the question of determining what's clever and what's true. As many students in the class came to observe, that's a question of authority as much as one of language. Interpreting words can be tricky.

Anyway, back to randomness. To get around the "to me" problem, we often try to look at language how it's actually used. For instance:

1. A guy comes up to you in the grocery store parking lot. He says, "I'm starting a business. If you loan me $100, I'll pay you back $200 next year." I think most people would think it appropriate to respond, "No offense, but I don't loan money to random people." Do you? But of course, on your definition of random as uniform probability distribution, it's not random at all. You only ran into this guy because you shopped at this store and not one in a different city or even across town; because you allowed him to approach you instead of walking away briskly; because you go to the store at all, etc.

2. The victims who were killed by the bullets meant for Trump in Butler PA. We call those deaths random, right? But they weren't random at all, per your definition. For one thing, you had to have been at a Trump rally to be shot. Second, you had to in the line of fire from the specific place where a gunman could get off a shot at Trump. Most people at the event had a 0% chance of being shot.

3. When people are asked to name a random number between 1 and 100, the results are not uniform. For one thing, almost nobody guesses 2, nor multiples of 10. I watched a video about this. I think the most chosen numbers (by a considerable margin, even) are 37 and 73. There's definitely a gradient.

But if I ask you to name a number between 1 and 100, any number, I think we'd describe that as random. And if I set the humidity in a commercial greenhouse to that number, my business partner might be justifiably outraged that I just set the growing conditions to a random number pulled out of someone's butt. Saying, "well, it's more likely to be 73 than 2" would not answer the criticism.

In general, I think that people use the word randomly to denote a process in which the outcome is not fully predictable, or if an outcome (or an input) is established by reasons that are not apparent.


Technical note: describing the probability distribution of a 2d6 isn't fully explained as "two separate random events." I could weight a 12 sided die so that its probability distribution would be the same as 2d6. You could say, well that simulates the 2d6 probability, which is true in a sense; but it's also true that the 2d6 actualizes a type of probability distribution that can be actualized in other ways as well. But that's not really important and also I might not be right about this point and I don't want to argue it.

Not "to me" as in personally to me, but "to me" as in as student of the fields of math and computer science where these words already have very precise technical meanings. I freely admit that jurists, lawyers, and legal scholars will freely either adopt or ignore these pre-existing meanings based on whim (or desired outcome, more likely).

 

[superrific]

No. Outcomes with uneven probability distribution are, by the very definition, decidedly not random. That's why it's precisely so damned hard to create a real random number generator, Almost every observable phenomena we encounter has some order built into it. Basically the best we can do is use cosmic background radiation as a proxy for true randomness.

Are you a computer programmer? Or use computer programming in your work? This strikes me as a definition situated within a technical context. It's not wrong per se (how can definitions be wrong?) but it's not useful because it's invoking a concept of concern to almost nobody. Outside of cryptography application, is the code x:=rnd insufficient to generate something we might call "random"? I've used that method of randomness many times and I've never had a problem with it. Monte Carlo simulations (random walks) are also random per the rnd test; if rnd is insufficient then a lot of scientific knowledge is actually false.

In addition, you've now invoked a different concept. The opposite of "order" is entropy, not randomness.

Here's a dictionary definition of random. Note that the definition your propose is definition 2b. So it's not completely wrong, but it's not the most common usage and the definition 2a is precisely what you are saying can't be random, by definition. Since we are trying to determine the ordinary meaning of the term, I'd say that definition doesn't help your case.

Definition of RANDOM

lacking a definite plan, purpose, or pattern; made, done, or chosen at random; relating to, having, or being elements or events with definite probability of occurrence… See the full definition

www.merriam-webster.com

 

[superrific]

Not "to me" as in personally to me, but "to me" as in as student of the fields of math and computer science where these words already have very precise technical meanings. I freely admit that jurists, lawyers, and legal scholars will freely either adopt or ignore these pre-existing meanings based on whim (or desired outcome, more likely).

Ah, this answered a question in my subsequent post. I figured you to be a programmer or involved in math or comp sci to some degree. And no offense, but this approach is very common among people with technical knowledge. They want their technical definitions to be the actual definitions in ordinary language, as if the relevant process of naming a phenomenon is denotative when in fact it's connotative.

To put the point differently, "random" is a word in the English language. My guess is that it was a word in English before probability theory was ever invented. By contrast, you have a mathematical principle that is actually defined in non-language terms (I don't know how to express randomness in pure math but I'm positive that it can be done). Then you attach that math concept to the term "random," because it's useful for you. That doesn't change the meaning of the word in English, any more than the use of the word compact in topology should mean that all cars can be described as "compact" since they are closed and bounded.

The law usually seeks to recover either the ordinary meaning of language, or sometimes the intended meaning of language. In interpreting words, we aren't ignoring pre-existing meanings so much as recognizing that words in English usually have multiple meanings, full of gradation.

 

[finesse]

No. Outcomes with uneven probability distribution are, by the very definition, decidedly not random. That's why it's precisely so damned hard to create a real random number generator, Almost every observable phenomena we encounter has some order built into it. Basically the best we can do is use cosmic background radiation as a proxy for true randomness.

I don't understand a bit of this so please treat this as ignorant and not stupid. I get lost where random requires a uniform probability distribution. How does that uniformity not affect the randomness? I have trouble reconciling that in my head.

 

[Ddseddse]

Are you a computer programmer? Or use computer programming in your work? This strikes me as a definition situated within a technical context. It's not wrong per se (how can definitions be wrong?) but it's not useful because it's invoking a concept of concern to almost nobody. Outside of cryptography application, is the code x:=rnd insufficient to generate something we might call "random"? I've used that method of randomness many times and I've never had a problem with it. Monte Carlo simulations (random walks) are also random per the rnd test; if rnd is insufficient then a lot of scientific knowledge is actually false.

In addition, you've now invoked a different concept. The opposite of "order" is entropy, not randomness.

Here's a dictionary definition of random. Note that the definition your propose is definition 2b. So it's not completely wrong, but it's not the most common usage and the definition 2a is precisely what you are saying can't be random, by definition. Since we are trying to determine the ordinary meaning of the term, I'd say that definition doesn't help your case.

Definition of RANDOM

lacking a definite plan, purpose, or pattern; made, done, or chosen at random; relating to, having, or being elements or events with definite probability of occurrence… See the full definition

www.merriam-webster.com

Almost every single rnd() function ever coded is correctly classified as a pseudo-random number generator. The correct use of the term "pseudo" here provides insight into what the precise meaning of random is (i.e. it's "pseudo" because it can't provide a perfectly flat distribution function).

But your example still proves MY point. Even though it is pseudo-random, it does it's very best to mimic the essential quality of randomness, which is a flat distributional function. If you want an unequal distribution functions you need to apply additional operation on top of rnd() (because even though it's pseudo-random, it's doing the best it can to provide an flat probability distribution.

If you use the "random" function rnd() it'd never going to do anything other than it's level best to give you a flat distribution function.

When we are sloppy with our language we sometimes increase the scope of the word "random" to include non-flat distribution functions (e.g. "randomly" flipping though a book), but that is a departure from the core meaning of the term.

Again, a jurist will chose or reject whatever definition based on the case they want to make (much as you are doing here, and hey, that's fine, that's the law profession), but none of that changes the essential nature of what randomness is.

ETA: I really don't have a stake in this fight. I'm mostly just arguing because it a) helps keep me sharp, b) is fun (I believe most lawyers share this belief) and c) I need an outlet for nervous energy. BTW, Super, I have an actual legal question over on the Jeff Jackson thread that I feel is right up your alley.

 

[Ddseddse]

I don't understand a bit of this so please treat this as ignorant and not stupid. I get lost where random requires a uniform probability distribution. How does that uniformity not affect the randomness? I have trouble reconciling that in my head.

Every outcome is equally likely. No output is more likely than another.

If one output is more likely than another then there is some order being imposed on the system externally, and it's by definition not random anymore.

 

[superrific]

Almost every single rnd() function ever coded is correctly classified as a pseudo-random number generator. The correct use of the term "pseudo" here provides insight into what the precise meaning of random is (i.e. it's "pseudo" because it can't provide a perfectly flat distribution function).

But your example still proves MY point. Even though it is pseudo-random, it does it's very best to mimic the essential quality of randomness, which is a flat distributional function. If you want an unequal distribution functions you need to apply additional operation on top of rnd() (because even though it's pseudo-random, it's doing the best it can to provide an flat probability distribution.

If you use the "random" function rnd() it'd never going to do anything other than it's level best to give you a flat distribution function.

When we are sloppy with our language we sometimes increase the scope of the word "random" to include non-flat distribution functions (e.g. "randomly" flipping though a book), but that is a departure from the core meaning of the term.

Again, a jurist will chose or reject whatever definition based on the case they want to make (much as you are doing here, and hey, that's fine, that's the law profession), but none of that changes the essential nature of what randomness is.

I think you're missing the point. That historian on the other board, who seems not to have made it here, would do the same thing.

The word random is not co-extensive with a computer scientist's definition of the word. Computer scientists coopted the word random and turned it into a word meaning uniform probability distribution. That's fine for their purposes, but it doesn't mean they now own the word random. They are using a word with broader meaning as a token to denote a more specific meaning.

The word random has its origins in Middle English and French, where it meant "at great speed." That connoted carelessness (I wonder if that comes from a military application), and then purposelessness. But for two centuries, the word has been used in English to mean unpredictable, not uniform probability distribution.

Thus have you inverted the causality. I'm not using some imprecise approximation of the word random. I'm using the word (or trying to) as it has been used in the language for centuries. Your logic here would produce odd results elsewhere:

1. "Work." In physics, work is force * distance. So moving boxes all day involves a lot of work. Seeing patients all day as a doctor actually involves little work at all. Even less work is required when lawyers stay up for 18 hours straight finalizing a merger or drafting a brief. Yet if you told that lawyer or doctor that they haven't done any word today, they would look at you like you were crazy. Or, depending on the extent of sleep deprivation, punch you.

2. "Obese" is defined by doctors as a certain body mass index. By BMI, I'm guessing Lebron James is obese. But if you called him obese, people would look at you like a crazy person. Likewise, most people who BMI considers obese are considered "overweight" in our culture, or not even that.

I don't have time for more examples. Hopefully you get the point.

 

[finesse]

Every outcome is equally likely. No output is more likely than another.

If one output is more likely than another then there is some order being imposed on the system externally, and it's by definition not random anymore.

Maybe this is the math/verbal divide because the stipulation of an equal likelihood seems like the imposition of some sort of order. I understand, I think, the math concept but it leaves my mental concept unsettled.

 

[Ddseddse]

Maybe this is the math/verbal divide because the stipulation of an equal likelihood seems like the imposition of some sort of order. I understand, I think, the math concept but it leaves my mental concept unsettled.

I get that. For me it helps to think of primeval chaos. Undifferentiated nothingness. No order or structure. What other result could there be but everything equally likely? As soon as any one thing becomes more likely than another, then you have imposed order upon on the primeval chaos.

 

[superrific]

ETA: I really don't have a stake in this fight. I'm mostly just arguing because it a) helps keep me sharp, b) is fun (I believe most lawyers share this belief) and c) I need an outlet for nervous energy. BTW, Super, I have an actual legal question over on the Jeff Jackson thread that I feel is right up your alley.

I don't know if I would call it "fun" but those reasons for continuing to discuss this are all valid.

 

[finesse]

I get that. For me it helps to think of primeval chaos. Undifferentiated nothingness. No order or structure. What other result could there be but everything equally likely? As soon as any one thing becomes more likely than another, then you have imposed order upon on the primeval chaos.

Chaos, to my mind, suggest anything but uniformity. It demands anything but, including for order to manifest but not stabilize. How, in a chaotic system, can a pattern even allow all chances be equal at one time. It almost insists that it be all chances are equal at some time.

 

[Ddseddse]

I think you're missing the point. That historian on the other board, who seems not to have made it here, would do the same thing.

The word random is not co-extensive with a computer scientist's definition of the word. Computer scientists coopted the word random and turned it into a word meaning uniform probability distribution. That's fine for their purposes, but it doesn't mean they now own the word random. They are using a word with broader meaning as a token to denote a more specific meaning.

The word random has its origins in Middle English and French, where it meant "at great speed." That connoted carelessness (I wonder if that comes from a military application), and then purposelessness. But for two centuries, the word has been used in English to mean unpredictable, not uniform probability distribution.

Thus have you inverted the causality. I'm not using some imprecise approximation of the word random. I'm using the word (or trying to) as it has been used in the language for centuries. Your logic here would produce odd results elsewhere:

1. "Work." In physics, work is force * distance. So moving boxes all day involves a lot of work. Seeing patients all day as a doctor actually involves little work at all. Even less work is required when lawyers stay up for 18 hours straight finalizing a merger or drafting a brief. Yet if you told that lawyer or doctor that they haven't done any word today, they would look at you like you were crazy. Or, depending on the extent of sleep deprivation, punch you.

2. "Obese" is defined by doctors as a certain body mass index. By BMI, I'm guessing Lebron James is obese. But if you called him obese, people would look at you like a crazy person. Likewise, most people who BMI considers obese are considered "overweight" in our culture, or not even that.

I don't have time for more examples. Hopefully you get the point.

Ah, unpredictable. You'll agree with me that a result can be more, or less, unpredictable, no? "Kamala Harris will win NC" (more unpredictable) vs. "Kamala Harris will win CA" (less unpredictable)... What discriminates one result as more unpredictable than another? The flattening of the probability distribution function, of course.

Squirm as you might, the flat probability function is at the core of randomness, there's no escaping that.

 

[Ddseddse]

Chaos, to my mind, suggest anything but uniformity. It demands anything but, including for order to manifest but not stabilize. How, in a chaotic system, can a pattern even allow all chances be equal at one time. It almost insists that it be all chances are equal at some time.

That's just lie told by Arioch, Lord of Chaos, to deceive his mortal followers. Primordial chaos is ultimately sterility and death

 

[finesse]

That's just lie told by Arioch, Lord of Chaos, to deceive his mortal followers. Primordial chaos is ultimately sterility and death

I have all those.

ETA: I do my homework now.

 

[superrific]

Ah, unpredictable. You'll agree with me that a result can be more, or less, unpredictable, no? "Kamala Harris will win NC" (more unpredictable) vs. "Kamala Harris will win CA" (less unpredictable)... What discriminates one result as more unpredictable than another? The flattening of the probability distribution function, of course.

Squirm as you might, the flat probability function is at the core of randomness, there's no escaping that.

1. Flattening of a probability function isn't the same thing as a completely flat probability function. That's just not how language works. Language is rarely taken to the logical extreme.

For instance, let's talk about the word "tall" as used to describe people. You can define tall operationally as the number of meters required to traverse a vertical plane starting at a person's feet and ending at their head (I know this isn't a plane per se, but I don't remember the right word. Surface?). But that doesn't mean that tall only refers to the tallest people. Indeed, to the contrary, tall usually has an implied reference to some sort of cutoff. You can put the cutoff in different places depending on your purposes -- maybe tall is someone whose body is longer than average? Or 80th percentile? 90th? -- but in the end, the word applies when the threshold is met. It doesn't continue further.

So too with random. That randomness is connected to flat probability distributions does not imply that only perfectly flat distributions are random, as the word is used in ordinary language. There's a threshold, and once crossed, the word applies.

Most adjectives I can think of work this way, because logical purity is not an important or valuable function of ordinary language. To the contrary, language is most useful when it's generalizable. If I apply for a job, and my resume is thrown out because HR rolls a 3d6 for each received resume and tosses any that come up 5 or less . . . well, I think I'd have a right to complain about that, and in formulating that complaint, I'd use the word random, as in "it's not fair to toss out resumes randomly like that." The function of the word is to contrast that selection process with something principled. I don't think anyone would dispute the use of the word there, nor in any similar way that might violate technical randomness as you've defined it.

In ordinary language, in fact, random often means "unimportant" or "arbitrary." For instance, if the employer tosses all resumes for people whose last name starts with the letter F, we'd call that exclusion "randomly." We'd say, "they have so many applicants that they can toss resumes out at random." That there was a method to the madness isn't really important, because the method is not considered acceptable.

2. I guess the main point here is that ordinary language and technical language have different functions in our society, and the desirable qualities of each are different. When you demand that ordinary language follow technical language -- there's just no justification for it. Computer scientists don't have special powers to change our language just because they define concepts with math.

If ordinary language borrows from technical language, the case for your position is stronger. For instance, DNA did not exist as a term until it was coined by biologists. So in this instance, biologists might be more justified in insisting that DNA refer to specific molecules in cells. But generally, they don't. People say, "doing the right thing is no longer in Boeing's DNA" even though obviously a company can't have DNA. Basketball analysts refer to players' gravitational pull -- meaning that their abilities, especially shooting abilities, require the defense to keep multiple players near them at all times. Obviously this isn't the same as gravity. And yet, many of those analysts are trained in math or physics and they don't object.

But in the usual case where technical definitions borrow from ordinary language, I just don't see how it can possibly be appropriate for a technical community to insist that its definition of the word has supplanted other usages as the correct definition. Are we going to need a new word for a small round dessert snack often filled with chocolate chips, because programmers defined "cookies" to refer to bits of information?

 

[Calheel]

Almost every single rnd() function ever coded is correctly classified as a pseudo-random number generator. The correct use of the term "pseudo" here provides insight into what the precise meaning of random is (i.e. it's "pseudo" because it can't provide a perfectly flat distribution function).

But your example still proves MY point. Even though it is pseudo-random, it does it's very best to mimic the essential quality of randomness, which is a flat distributional function. If you want an unequal distribution functions you need to apply additional operation on top of rnd() (because even though it's pseudo-random, it's doing the best it can to provide an flat probability distribution.

If you use the "random" function rnd() it'd never going to do anything other than it's level best to give you a flat distribution function.

When we are sloppy with our language we sometimes increase the scope of the word "random" to include non-flat distribution functions (e.g. "randomly" flipping though a book), but that is a departure from the core meaning of the term.

Again, a jurist will chose or reject whatever definition based on the case they want to make (much as you are doing here, and hey, that's fine, that's the law profession), but none of that changes the essential nature of what randomness is.

ETA: I really don't have a stake in this fight. I'm mostly just arguing because it a) helps keep me sharp, b) is fun (I believe most lawyers share this belief) and c) I need an outlet for nervous energy. BTW, Super, I have an actual legal question over on the Jeff Jackson thread that I feel is right up your alley.

But is anything in the universe truly random? Universal causation would suggest not.

But none of these philosophical questions should be determinative of Elon's stupid argument about the lottery.