Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Programming Language Creator or Serial Killer?

vole.wtf

Post reply on HN
Print view View original

131–140 of 148 posts

Re: Programming Language Creator or Serial Killer?

#131

Post by Geminidog »

gus_massa wrote: post #121
Geminidog wrote: post #96

Earlier quoted context omitted.

I got 8/10 which says otherwise. I voted serial killer for both women just because of how unlikely it is for a female to invent a programming language. If it wasn't for that I would be basically 90% correct. This means statistically that there likely is a difference. I can't put my finger on what it was though, I just knew. Basically anybody who looked insane must be an inventor of a programming language and anyone w…

The probability to get 8/10 choosing randomly with equal probability is 4.4%. This has 241 points now, so if we assume that the number of people that took the test and has an account to write a comment is similar to the amount of people that upvoted, we get 10.6 with 8/10. I prefer to round it to 10, because there are too many approximations. So it's not unexpected to get a few person that got 8/10, even if it were a…

The logic is sound when applied to my singular observation. But a good number of people got > 8/10. Read the thread.

Re: Programming Language Creator or Serial Killer?

#132

Post by Blikkentrekker »

antihero wrote: post #120
Blikkentrekker wrote: post #115

Earlier quoted context omitted.

I find that, ironically, a great deal of advice given to people on how to “be safe” is very counter-effective, for it inspires within them the mistaken confidence that they can easily recognize who is dangerous, and are thus more likely to drop their guard when such recognition yields a false negative. I also recently saw a comment on a forum along the lines of “ If Hollywood started to consistently cast beautiful pe…

Well that's the issue, if we give people rules for people to follow when assessing the danger of a situation or someone's behaviour, then they are at a danger to anyone who can hack those rules and not set off the red flags. We have to teach people intuition, and to deeply question people's motives and what they have to gain versus what the person has to lose.

One doesn't even have to deliberately hack them; acting as one naturally does will be sufficient.

Most human beings do not naturally act as easily-identifiable Hollywood villains.

Re: Programming Language Creator or Serial Killer?

#133

Post by gus_massa »

Geminidog wrote: post #131
gus_massa wrote: post #121

Earlier quoted context omitted.

The probability to get 8/10 choosing randomly with equal probability is 4.4%. This has 241 points now, so if we assume that the number of people that took the test and has an account to write a comment is similar to the amount of people that upvoted, we get 10.6 with 8/10. I prefer to round it to 10, because there are too many approximations. So it's not unexpected to get a few person that got 8/10, even if it were a…

The logic is sound when applied to my singular observation. But a good number of people got > 8/10. Read the thread.

I count 6 people with 8/10. So there are approximately 4 shy ones.

Anyway, one said that can recognize Guido. If you know one, you only have to get 7 out of 9, that has a probability of 7%. As expected, knowing one answer make the test easier.

Re: Programming Language Creator or Serial Killer?

#134

Post by Geminidog »

gus_massa wrote: post #133
Geminidog wrote: post #131

Earlier quoted context omitted.

The logic is sound when applied to my singular observation. But a good number of people got > 8/10. Read the thread.

I count 6 people with 8/10. So there are approximately 4 shy ones. Anyway, one said that can recognize Guido. If you know one, you only have to get 7 out of 9, that has a probability of 7%. As expected, knowing one answer make the test easier.

The numbers are off. You can't have votes be your basis in how many people took the test ( I didn't vote for example) and you can't have the amount of 8/10 comments be a metric for the sole count of people who scored that much. The numerical game can't be played here because the information is too cloudy.

That being said, anecdotally a good number of people got 8/10, 9/10 and 10/10 which anecdotally makes it likely to be statistically significant. Hard to say quantitatively but we can make a qualitative prediction here.

Re: Programming Language Creator or Serial Killer?

#135

Post by tooltower »

mattround wrote: post #70
tooltower wrote: post #8

Why do people still think they can judge a person's character by a single photograph? This is an age-old dogma rooted in genetic superiority crap that really needs to die. Even a simple Google search shows a staggering number of these pop-psychology articles being actively peddled to the masses. Oh, if a person's nose is a bit crooked in the wrong direction, watch out!

Hi, I made the quiz and tried to design it to make judging by appearance work as poorly as possible, so the initial daft premise then becomes something a little bit more interesting as you get some wrong. But I think it's reasonable to be wary of the issue, particularly with endless problems emerging nowadays with facial recognition tech.

Thanks for the response. I do appreciate the intent. As you can see, though, some of us don't are more prone to missing the intent. It's even demonstrated by one of the responses to my comment here.

Maybe make that end goal a bit more explicit towards the end? Or maybe it's okay as is, since it did generate a conversation and thus served its purpose.

Out of curiousity, how much work (e.g. in days or weeks) does it take for you to make something like this?

Re: Programming Language Creator or Serial Killer?

#136

Post by gus_massa »

Geminidog wrote: post #134
gus_massa wrote: post #133

Earlier quoted context omitted.

I count 6 people with 8/10. So there are approximately 4 shy ones. Anyway, one said that can recognize Guido. If you know one, you only have to get 7 out of 9, that has a probability of 7%. As expected, knowing one answer make the test easier.

The numbers are off. You can't have votes be your basis in how many people took the test ( I didn't vote for example) and you can't have the amount of 8/10 comments be a metric for the sole count of people who scored that much. The numerical game can't be played here because the information is too cloudy. That being said, anecdotally a good number of people got 8/10, 9/10 and 10/10 which anecdotally makes it likely t…

I think that everyone that upvoted played the game. So I think that at least 241 persons played the game. (The actual number of players is probably way higher.)

If you assume that players have no knowledge, With 241 the expected to number of persons with 8/10 in 10, and with 10/10 is 0.2. (The actual number of players is probably way higher, so these expected values are way higher, but we must not count lurkers.)

Last time I counted, I found 6 comments that say 8/10, so even if only 241 persons played and none of them have any insight, it would be underreported. This is not unexpected, not everyone is posting the result.

Last time I counted, I found 3 comments that say 10/10, that's much more than the expected value if only 241 persons played and none of them have any insight. So that's slightly interesting. (Did a lot of people get 10/10 and didn't upvote? The people that say 10/10 is lying or cheating? Do they recognize a few of the photos and were lucky in the others? Do they have a sixth sense to recognize serial killers?...)

Re: Programming Language Creator or Serial Killer?

#137

Post by Geminidog »

dr_kiszonka wrote: post #104
Geminidog wrote: post #96

Earlier quoted context omitted.

I got 8/10 which says otherwise. I voted serial killer for both women just because of how unlikely it is for a female to invent a programming language. If it wasn't for that I would be basically 90% correct. This means statistically that there likely is a difference. I can't put my finger on what it was though, I just knew. Basically anybody who looked insane must be an inventor of a programming language and anyone w…

With a sample of one, it is hard to make any meaningful inferences ;) But intuitively, I do tend to agree with your point. Perhaps PL creators look a touch more... extravagant?

But I took 10 samples.

Let's take it to the extreme. Lets say the test had 100 questions.

If out of all the people who took the test, one person got a score of 100/100 out of 10 million people who took the test it means that there are meaningful differences, it's just that most people can't identify those differences.

Re: Programming Language Creator or Serial Killer?

#138

Post by salamanderman »

Geminidog wrote: post #126
salamanderman wrote: post #122

Earlier quoted context omitted.

I assume you were trying to describe how you did well using statistics and demographics, but your wording comes off sexist. It hits my ear as bragging about the disparity in the field, and I have to consciously try to change my initial interpretation to allow for something different. I also got an 8/10, but I had at least one woman in my set who was a language developer (one of the Cobol designers) and I got that rig…

I said something akin to men are unlikely to be nurses, so I didn't choose men. I think the ideological rhetoric of feminism has infected the viewpoints of normal people so much so that one needs to tiptoe around something that is obviously true. This may be good from a social standpoint as it opens up equality of opportunity but it is bad for science. The movement has silenced people who point out inevitable differe…

I'm sorry I gave you the benefit of the doubt. I shouldn't take the bait but, hey, you played a GAME that was structured such that you'd have a 50/50 shot of being right. Rather than assume that the balance of demographics in the photos were waited towards that game's premise, 50/50 split, you used your own biased heuristic and happened to get a good score. You happened to get a set of 10 that included no female language developers. I had a mixed set. 8 coin flips out of 10 of heads is not special. I had a heuristic too and I'm probably equally wrong about it meaning something. I'm being honest with myself, though.

Re: Programming Language Creator or Serial Killer?

#139

Post by yakubin »

bryanrasmussen wrote: post #111
Geminidog wrote: post #96

Earlier quoted context omitted.

I got 8/10 which says otherwise. I voted serial killer for both women just because of how unlikely it is for a female to invent a programming language. If it wasn't for that I would be basically 90% correct. This means statistically that there likely is a difference. I can't put my finger on what it was though, I just knew. Basically anybody who looked insane must be an inventor of a programming language and anyone w…

it's also statistically unlikely for females to be serial killers as well.

It's also statistically unlikely for males to be serial killers as well.

Re: Programming Language Creator or Serial Killer?

#140

Post by Geminidog »

salamanderman wrote: post #138
Geminidog wrote: post #126

Earlier quoted context omitted.

I said something akin to men are unlikely to be nurses, so I didn't choose men. I think the ideological rhetoric of feminism has infected the viewpoints of normal people so much so that one needs to tiptoe around something that is obviously true. This may be good from a social standpoint as it opens up equality of opportunity but it is bad for science. The movement has silenced people who point out inevitable differe…

I'm sorry I gave you the benefit of the doubt. I shouldn't take the bait but, hey, you played a GAME that was structured such that you'd have a 50/50 shot of being right. Rather than assume that the balance of demographics in the photos were waited towards that game's premise, 50/50 split, you used your own biased heuristic and happened to get a good score. You happened to get a set of 10 that included no female lang…

>I'm sorry I gave you the benefit of the doubt.

This is basically your way of saying that my viewpoint is sexist. I have not acted in such a manner, and please cite evidence as to where I am acting that way if you believe otherwise. Basically all I'm saying is that people should stop turning every freaking conversation about real demographic differences into an ideological witch hunt. It's completely off topic and nobody is thinking about such things yet there always has to be some social justice warrior pointing out some "infraction" whoops did I say women are physically weaker than men even though it's generally the truth? My bad for being a heathen sexist.

>Rather than assume that the balance of demographics in the photos were waited towards that game's premise, 50/50 split.

I assumed the balance was not weighted towards sex. I assumed he picked random set of people where the sex and race were random and only whether or not they were language developers/serial killers was the heuristic. It was a mistaken assumption because if the picks were truly random, it is highly unlikely for women to appear in the question set at all (assuming women are unlikely to be serial killers as stated by some people in this thread). Obviously, because two women appeared in the question set, the heuristic is biased for other attributes as the test maker deliberately made sure the question set included two women.

Frankly, the test actually masks reality as we know it. Programmers, Language creators and serial killers are highly, highly unlikely to be women.

Post reply on HN