Eddie has some math talent. He can invest some time, money, and effort C to get a degree, which allows other people to discern that he has a higher probability of having that math talent. This higher probability confers some benefit in that other people will more readily take his advice in mathematical matters, or talk with him about his math.
The fun twist is that Eddie lives in a society with many other individuals with varying degrees of math talent, each of whom can expend C to get a degree and the associated benefits. People with almost no mathematical talent have a prohibitively high C, because even if they can pony up the time and money, they have to work very hard to fake their way through. But people with high math ability often choose to stand out by getting the degree, because their C is relatively lower, and a very high proportion of them get degrees. This creates a high association between degrees and mathematical ability, and makes it unlikely to see high mathematical ability in the absence of a degree.
That’s the basic idea, plus degrees signal other things which may be completely unrelated to math, but are still nice. Even in the case where the degree has no causal effect no math ability, there are benefits to having one, in that the other math people can judge very quickly that they’re interested in talking to you.
Hopefully that demonstrates that I understand signalling. My question is about the costs and benefits of a particular signal.
It demonstrate that you don’t. Humans make decisions via something called the availability heuristic.
If you bring into the awareness of the person that you are talking that you are a mathematician that only has a bachleor, no master, no PHD and no professorship that you aren’t bringing expertise into his mind.
If you are however a self taught person who managed to published multiple papers among them a paper titled “Complex Value Systems in Friendly AI” in Artificial General Intelligence Lecture Notes in Computer Science Volume and who has his own research institute that’s a better picture.
If you have published papers that a lot more relevant for relevant experts than whether you have a degree that verifies basic understanding.
If a person really cares whether Eliezer has a math degree he already lost that person.
Eddie has some math talent. He can invest some time, money, and effort C to get a degree, which allows other people to discern that he has a higher probability of having that math talent. This higher probability confers some benefit in that other people will more readily take his advice in mathematical matters, or talk with him about his math.
The fun twist is that Eddie lives in a society with many other individuals with varying degrees of math talent, each of whom can expend C to get a degree and the associated benefits. People with almost no mathematical talent have a prohibitively high C, because even if they can pony up the time and money, they have to work very hard to fake their way through. But people with high math ability often choose to stand out by getting the degree, because their C is relatively lower, and a very high proportion of them get degrees. This creates a high association between degrees and mathematical ability, and makes it unlikely to see high mathematical ability in the absence of a degree.
That’s the basic idea, plus degrees signal other things which may be completely unrelated to math, but are still nice. Even in the case where the degree has no causal effect no math ability, there are benefits to having one, in that the other math people can judge very quickly that they’re interested in talking to you.
Hopefully that demonstrates that I understand signalling. My question is about the costs and benefits of a particular signal.
It demonstrate that you don’t. Humans make decisions via something called the availability heuristic.
If you bring into the awareness of the person that you are talking that you are a mathematician that only has a bachleor, no master, no PHD and no professorship that you aren’t bringing expertise into his mind.
If you are however a self taught person who managed to published multiple papers among them a paper titled “Complex Value Systems in Friendly AI” in Artificial General Intelligence Lecture Notes in Computer Science Volume and who has his own research institute that’s a better picture.
If you have published papers that a lot more relevant for relevant experts than whether you have a degree that verifies basic understanding. If a person really cares whether Eliezer has a math degree he already lost that person.