Not defending musk, but it is really difficult to find women in AI. The few that are around are very valuable for tech companies that will pay higher salaries to have “token women”, better if they are from some minority group, to show better diversity statistics. Therefore retention is also difficult. Finding good people in AI is already difficult and expansive, finding women is a real challenge.
I strongly believe that diversity brings a lot of value, and women are important in any team. But the solution unfortunately is not in the current market, but it is at school levels. Culture must change.
Retention is indeed a problem. I don’t think I qualify as a “woman in AI,” but I am a cis woman who has trained (well, fine-tuned) my own models on my gaming PC at home as a hobby. Several years ago, I fucked off and became a professional photographer after working for a Fortune 50 for a decade; I loved my job but hated the sexism. There’s almost no amount of money that could get me back to working in tech.
(Incidentally, a bunch of my images were scraped and used in the training data set for Stable Diffusion. I’m mad about this and have no desire to help corporations profit off others’ art.)
From my experience ML and data science in general are very welcoming to women and people with very different backgrounds. Also the way of working is very different. Agile doesn’t really work, because is a non-deterministic world, you have relatively “long” projects, no PM chasing burndown (burnout) story points (or whatever those silly metrics are called), curious and interesting colleagues that are there for passion. You can give it a try. As said, in many industries, women in data science and ML are highly valued and unfortunately there are not enough of them.
If women only see men or almost only men in jobs they’re going to naturally assume they aren’t wanted in those fields. And let’s face it, as much as diversity is sought after women AREN’T wanted in tech by a really large portion of the gatekeeping dudes pulling the strings.
Like, sure, I believe you about the hiring challenges but this is an Elon Musk company - do you even think he tried? Did you see what happened to Twitter’s personnel diversity when he arrived?
He can use it as a defense, but trust me, I am not defending him.
What I am trying to say is that in some industries the inequality is in the market. Whoever manage to balance it is doing it by putting down a lot of money and providing extra benefits (such as quick career progression). This means that whoever is not ready to do so will end up with a very imbalanced distribution. Musk has demonstrated that his focus is saving money, which is probably also one of the reasons he is ending up with only men. There are for sure others, but I believe it makes sense to consider the overall market.
To “overcome” the imbalance, you cannot really do it on the current market, because it is already imbalanced. What you have is companies that fight for a small pool of available employees, some will meet a decent diversity (with effort and money) others won’t be able to do so.
The problem has to be solved before, at a cultural level. As I said in another comment, for instance this imbalance is not a problem in pharma, where data science is probably majority women (I don’t know the stats, but this is my anecdotal experience). There is a cultural issue with women in tech that can be solved only with cultural changes and with generational change. Already z gen is in a better position than millennials or x gen. But more needs to be done
Saying “coincidence” is basically claiming there is no reason for an observed pattern. This is really more of a last resort when considering explanations for certain patterns, because it’s probably the weakest claim someone can make.
Generally, patterns are not coincidence because if an outcome is truly a result of randomness, then there is an extremely low chance that there would be a pattern.
Also, 12 is not the whole data set. The whole data set should include the people who weren’t hired during the hiring process. This is unknown to us.
Nice stats, but it isn’t broken down on industry. From experience (I worked in different fields) in some industries such as pharma, people analytics or marketing, women are even likely the majority (they were majority when I worked in pharma, for instance). In more “pure” tech and fintech companies, I do not believe those stats represent the “natural distribution”. I know it’s anecdotal, but trust me, it’s not easy to find woman in AI in some industries. They are highly valued, well paid and have quick career progression because of this, to attract and retain them.
That said, elon is probably “machist” type of guy, I am not defending him. Just trying to give a bit of context
The entire paper is already sub-field (AI) in industry (software engineering) specific. No stats are perfect, but I think these ones are pretty damn good for something where peoples role are pretty poorly determined in the first place. Of course you’re welcome to try and find better ones.
The “pure tech” companies I’ve worked at have been roughly equivalent or better than these stats, but at that point I’m sampling from software engineers in general (not having worked at an AI specific company), and my sample is unlikely to be unbiased anyways.
AI is in all industries, from pharma, insurance, finance. Nowadays it is not even a sub field of software engineering, more of a subfield of data science. If you check the background of those who work in AI, you find the most varied combinations, from maths, to engineering, stats, and physics.
I don’t have better statistics unfortunately. And I don’t even want to be right.
My anecdotal experience is that women cluster in some industries, in other industries they are difficult to find, in AI more than in other subfields of data science such as what is historically defined as “statistical inference”.
Again this is anecdotal, not hard science. But, as we don’t have stats, better than nothing.
Edit. Again, not trying to defend anyone, just adding information for people to draw their own conclusions.
Mine are that elon was trying to save some money, and he doesn’t value diversity to invest on it, and put the extra effort to create it
Taking 89.3% men from your source at face value, and selecting 12 people at random, that gives a 12.2% chance (1 in 8) that the company of that size would be all male.
Add in network effects, risk tolerance for startups, and the hiring practices of larger companies, and that number likely gets even larger.
What’s the p-value for a news story? Unless this is some trend from other companies run by Musk, there doesn’t seem to be anything newsworthy here.
Not defending musk, but it is really difficult to find women in AI. The few that are around are very valuable for tech companies that will pay higher salaries to have “token women”, better if they are from some minority group, to show better diversity statistics. Therefore retention is also difficult. Finding good people in AI is already difficult and expansive, finding women is a real challenge.
I strongly believe that diversity brings a lot of value, and women are important in any team. But the solution unfortunately is not in the current market, but it is at school levels. Culture must change.
Retention is indeed a problem. I don’t think I qualify as a “woman in AI,” but I am a cis woman who has trained (well, fine-tuned) my own models on my gaming PC at home as a hobby. Several years ago, I fucked off and became a professional photographer after working for a Fortune 50 for a decade; I loved my job but hated the sexism. There’s almost no amount of money that could get me back to working in tech.
(Incidentally, a bunch of my images were scraped and used in the training data set for Stable Diffusion. I’m mad about this and have no desire to help corporations profit off others’ art.)
From my experience ML and data science in general are very welcoming to women and people with very different backgrounds. Also the way of working is very different. Agile doesn’t really work, because is a non-deterministic world, you have relatively “long” projects, no PM chasing burndown (burnout) story points (or whatever those silly metrics are called), curious and interesting colleagues that are there for passion. You can give it a try. As said, in many industries, women in data science and ML are highly valued and unfortunately there are not enough of them.
If women only see men or almost only men in jobs they’re going to naturally assume they aren’t wanted in those fields. And let’s face it, as much as diversity is sought after women AREN’T wanted in tech by a really large portion of the gatekeeping dudes pulling the strings.
Like, sure, I believe you about the hiring challenges but this is an Elon Musk company - do you even think he tried? Did you see what happened to Twitter’s personnel diversity when he arrived?
As said, I am not trying to defend him. He behaved extremely unprofessionally at Twitter.
It was just to give a bit of context
But it is a defense. Every unjustifiable hierarchy or segregation has been at one point deemed too difficult to overcome to be worth it.
He can use it as a defense, but trust me, I am not defending him.
What I am trying to say is that in some industries the inequality is in the market. Whoever manage to balance it is doing it by putting down a lot of money and providing extra benefits (such as quick career progression). This means that whoever is not ready to do so will end up with a very imbalanced distribution. Musk has demonstrated that his focus is saving money, which is probably also one of the reasons he is ending up with only men. There are for sure others, but I believe it makes sense to consider the overall market.
To “overcome” the imbalance, you cannot really do it on the current market, because it is already imbalanced. What you have is companies that fight for a small pool of available employees, some will meet a decent diversity (with effort and money) others won’t be able to do so.
The problem has to be solved before, at a cultural level. As I said in another comment, for instance this imbalance is not a problem in pharma, where data science is probably majority women (I don’t know the stats, but this is my anecdotal experience). There is a cultural issue with women in tech that can be solved only with cultural changes and with generational change. Already z gen is in a better position than millennials or x gen. But more needs to be done
Eh, the gender imbalance is bad, but not 0/12 bad… here are some stats
deleted by creator
Saying “coincidence” is basically claiming there is no reason for an observed pattern. This is really more of a last resort when considering explanations for certain patterns, because it’s probably the weakest claim someone can make.
Generally, patterns are not coincidence because if an outcome is truly a result of randomness, then there is an extremely low chance that there would be a pattern.
Also, 12 is not the whole data set. The whole data set should include the people who weren’t hired during the hiring process. This is unknown to us.
deleted by creator
Those stats don’t take into account the number of women that would want to work for Elon Musk.
Isn’t the fact that he’s repulsive sort of the whole complaint?
Just saying… if you take that variable into account it probably gets a lot closer to that 0/12
Nice stats, but it isn’t broken down on industry. From experience (I worked in different fields) in some industries such as pharma, people analytics or marketing, women are even likely the majority (they were majority when I worked in pharma, for instance). In more “pure” tech and fintech companies, I do not believe those stats represent the “natural distribution”. I know it’s anecdotal, but trust me, it’s not easy to find woman in AI in some industries. They are highly valued, well paid and have quick career progression because of this, to attract and retain them.
That said, elon is probably “machist” type of guy, I am not defending him. Just trying to give a bit of context
The entire paper is already sub-field (AI) in industry (software engineering) specific. No stats are perfect, but I think these ones are pretty damn good for something where peoples role are pretty poorly determined in the first place. Of course you’re welcome to try and find better ones.
The “pure tech” companies I’ve worked at have been roughly equivalent or better than these stats, but at that point I’m sampling from software engineers in general (not having worked at an AI specific company), and my sample is unlikely to be unbiased anyways.
AI is in all industries, from pharma, insurance, finance. Nowadays it is not even a sub field of software engineering, more of a subfield of data science. If you check the background of those who work in AI, you find the most varied combinations, from maths, to engineering, stats, and physics.
I don’t have better statistics unfortunately. And I don’t even want to be right.
My anecdotal experience is that women cluster in some industries, in other industries they are difficult to find, in AI more than in other subfields of data science such as what is historically defined as “statistical inference”.
Again this is anecdotal, not hard science. But, as we don’t have stats, better than nothing.
Edit. Again, not trying to defend anyone, just adding information for people to draw their own conclusions.
Mine are that elon was trying to save some money, and he doesn’t value diversity to invest on it, and put the extra effort to create it
“my anecdotal industry experience trumps your stats” you don’t sound like you have a very unbiased opinion brah
If you read again, you’ll see I am saying that the stat is not complete as it doesn’t drill down to industries brah.
In absence of statistics anecdotal evidence is better than nothing to draw qualitative conclusions brah
If you have different experience, I am happy to discuss
deleted by creator
Taking 89.3% men from your source at face value, and selecting 12 people at random, that gives a 12.2% chance (1 in 8) that the company of that size would be all male.
Add in network effects, risk tolerance for startups, and the hiring practices of larger companies, and that number likely gets even larger.
What’s the p-value for a news story? Unless this is some trend from other companies run by Musk, there doesn’t seem to be anything newsworthy here.