I don't see an issue with InfoWars having a platform or even to be on the same platform as me but I'd have serious reservations about having them as the platform manager, mainly because I'm not trusting them to have sane opinions (it's not like InfoWars strike me as vanguards of free speech in principle).
Lol, apparently the water turned frogs like me gay? I wouldn't trust InfoWars with my supporter's cash, Patreon, Kickstarter & ilk are in a position that demands a high level of trust, seeing as they deal with large quantities of money and the relationships that surround them.
I think the question you should be asking is why are people taking these "too low" wage jobs?
Companies do close down all the time because expenses are higher than revenues. Restaurants are a very well-known industry where that is super common (and it's not because restaurants tend to pay exhorbitant wages). It's literally a meme to accountants that having a restaurant as a client is a good way to never get paid (they may close down in the 30 days after delivery they are typically given to pay you).
Also, in a way, we already "shut down" companies that pay too little: you just can't even offer to pay someone 2$/hour. The question I'd ask is "who would take that job if it were allowed?"
In essence, having low wage jobs should never be an issue: if you're of the opinion that these companies shouldn't exist in the first place, you're essentially saying that those workers have better options already available. Do we really need to fire them to get them to go for those?
Yes? Or just get hired by one? Or change field? Or move?
It's no fable. Some of us don't have what it takes to start their own company, that's true.
Probably there is no hope for some. I don't know how you'd show that there is a system that doesn't suffer from this problem. I mean, you'd have to show that even with less incentive to produce and more demand for products (since everyone is owed "decent living" or whatever it's called nowadays), there is no possibility for demand to exceed production.
One question that would be interesting to explore is: what is the predictive score of judges/juries/etc? Is it more or less variable than the COMPAS score?
In other words, do judges perform better than COMPAS? And slightly less important: is the variance lower?
If it isn't, then there is no point arguing whether COMPAS is risky: it would be less so than the alternative.
Although I suppose that the 'crowd of non-experts' is essentially what a jury is and they had about the same performance.
At any rate, it sounds like much less of a hassle to input characteristics into COMPAS and get a probability than getting a jury together.
Besides, I think I'd be much more worried that a jury could influence each other and bias its decision than about am algorithm making unfair predictions.
Because it is a detail virtually nobody cares about.
How insane would the password rules have to be for anyone to travel 1 hour more to go to a different university? How insane for them to stop playing a given video game? To change banking institutions?
I don't have the answer for others, but for me, the answer to all of those is "pretty insane". Except for the banking case, password security is a minor concern (and even then, the system protects us with anti-fraud laws and what not).
In a sense, the market is sorting itself out: it just decided that it doesn't care much about passwords. In fact, if you figure out a way to be profitable while offering twice the interest rate but every time people log in to your bank they have to dance the robot or whatever, you'd probably still have customers.
I think you have to just assume that there is a lot of stupid in any system, so it can't be _all_ of those questions.
The people out there who drive drunk or without insurance aren't doing it with the intent of killing people or not having to pay for their damages: they're doing it because they "didn't think about it".
Hanlon's razor tells us that a similar phenomenon is at play in many "bad interview questions".
Or, the emergent result of our hiring processes (alienating and near-untenable interviews coupled with fast tracks for sponsored candidates) is that cultural outsiders (women, people of color, older workers) are effectively locked out of higher-status resume-building jobs, and when companies try to fix those processes even a little, engineers are confronted with culturally unfamiliar potential coworkers and experience the human default of hypervigilance in unfamiliar social situations, which creates a false perception that reformed, rationalized hiring processes are "lowering the bar" when in fact the opposite would probably be happening, with a long term net result that nothing changes, and our industry remains structurally biased against people who can't pass for proto-Zuckerbergs.
I call affirmative action policies lowering the bar. If you are intentionally seeking people based on criteria like race or gender, then you are less likely to find the best candidates, unless you believe those are relevant criteria.
What does intention have to do with it? Are you saying that if you unintentionally seek people based on criteria like race or gender you're more likely to find the best candidates? Affirmative action, as you call it, is an intentional effort to overcome the sort of unintentional bias that tptacek described.
By the way, one reason people react strongly to objections like yours is that a very direct interpretation is that you believe the distribution of people in high status jobs actually does accurately reflect differences based on race and gender. Not sure if you intended that.
> By the way, one reason people react strongly to objections like yours is that a very direct interpretation is that you believe the distribution of people in high status jobs actually does accurately reflect differences based on race and gender.
I think this is partly true. But then, a similar interpretation of "pro-affirmative action" is that one believes that whatever bad thing happens to a minority is never their fault.
I think both are only partly true.
The problem with affirmative action is that it (I don't know if it's always the case, but it seems to be often so) specifically considers minority status. It is obvious, I think, that this is not the kind of criteria a computer, say, would consider if it were to look only for the most qualified applicants.
On the other hand, you or someone else in the thread made a good point: this pro-minority bias can be seen as an effort to counter the systematic bias against "outgroups".
The problem is that I don't think we know the effect size of each of those: how strong is this systematic bias? In a protected status agnostic society, would all professional fields perfectly mirror population rates?
I don't think so. We know there are metal disorders/illnesses that impact cognitive functions (Down's syndrome) and obviously genetics is partly responsible (do Down syndromers have similar rates of Down syndrome babies?). Similarly, I don't think the success of Asians and Jews is due to some form of societal bias in their favor.
In the end, the question is whether the anti-bias bias is correcting towards 0 or in the other direction (i.e.: protected status is very important, just the other way).
False. If you believe that the best candidates are spread across genders/races and your current hiring practices are overlooking those not in a specific set, affirmative action policies would be raising the bar. You call it lowering the bar because you believe that all the best candidates are already being hired under the current system.
You misread my comment, and in fact took away kind of the opposite intervention and perceived response to the intervention that I was referring to.
I don't blame you for not following, because I was sort of glibly/ironically summarizing a lot of other things I've written about[1] as a single very complicated sentence. However, I sort of do blame you for jumping at the opportunity to litigate affirmative action on this thread.
If you re-read my comment and come to understand what I mean by this, would you please take a moment and write a short comment to acknowledge that? Thanks.
> I call affirmative action policies lowering the bar.
To turn around the vocabulary a little, isn't the fact that most jobs in SV are effectively (rarely in a formalized sense and sometimes even unconsciously) 'reserved' for white and Asian males - isn't that the overwhelming majority of 'affirmative action' cases?
Very likely, many white and Asian guys who get hired are not the best candidates. Baseball is a simpler example because the number of jobs is strictly defined (25 per team, N number of teams): Before it was integrated, starting with Jackie Robinson in 1947, a lot of minor-league-level white guys had major league jobs for which they were unqualified, and it was because of what was effectively a huge 'affirmative action' program for white players. Imagine if for some horrible reason baseball re-segregated today, eliminating all non-white players (Latinos were banned before 1947 too): Who would take the jobs of all the qualified non-white major leaguers? White-skinned minor leaguers. Beyond a doubt, the quality of baseball would suffer greatly - which, if you think about it, seems very likely to be true of the quality of SV companies' talent pools.
If you care about merit, then the current system is a failure.
> If you are intentionally seeking people based on criteria like race or gender, then you are less likely to find the best candidates
Agreed, but evidently that is what is happening already on an overwhelming basis, except it is white-skinned and Asian people who get special treatment.
Real-world hiring isn't so meritocratic and discrimination has always overwhelmed merit (again, look at baseball, or all hiring processes back then - the same happens now). Beyond merit, how are people really selected?: 1) People hire those they know or are in their network, and white people tend to network with other white people (because that is who they work with and went to school with, due to past discrimination -- it's self-reinforcing). 2) People are prejudiced, some overtly, more covertly, and very many without realizing it. 3) People hire to not fail - the decision-maker doesn't want to stick their neck out and bring in someone unpopular; they want to fit in with their co-workers and not be the social crusader. 4) People tend to hire others who are like themselves; when someone told me about this tendancy, I realized I'd unconsciously been doing it for years.
Affirmative action-like programs balances those forces a little (not much, looking at outcomes).
> the title implies that heritage impacts cognitive ability.
In this case, the title was misleading, but genetic heritage does impact cognitive ability, otherwise we'd have no cognitive advantage over our Great Ape distant cousins.
Not-so-minor correction: the message scrawled on the asphalt in the last photo says "Faure doit partir", which is French for "Faure MUST leave", not "should".
> The practical difference is that I think there should be further research to design inherently safe approaches and processes to reduce the risk until it's orders of magnitude safer
In practice this attitude makes you pro-coal (but not in a dogmatic way).
In the real world, waiting for things to improve before pulling the trigger means using what's currently used until then. That has costs.
What's the energy source that's most dangerous? The common sense answer is nuclear (probably because of nuclear bombs first, Chernobyl second). Some clever people might argue about coal given its high emission output.
But at least by death count, the winner is clearly hydro power. Mostly due to the Banqiao Dam incident [1] which directly killed 26,000 people and indirectly killed 145,000.
Of course, that incident is a mess of stupid decision after stupid decision and some might want to claim that it's an outlier. To that I will reply that the same can be said of Chernobyl, Three Mile Island and Fukushima (as well as every other hydro incident[2]).
The fact is that the world is full of stupid. I predict that in the future there will be some solar power company that figures out how to kill a bunch of people with a battery fire. And yet, that won't change anything to my support of hydropower or solar power. Nuclear safety is a minor issue compared to its advantages over coal and other fossil fuels.
I know nothing about Visual Studio Code and very little about anything VS in fact, so I have no idea how much is VS's fault and how much is the user's, but seriously, how on Earth is it possible to have 3 months of work not backed up or source controlled in any way?
I once had a nasty file system corruption due to disconnecting the power cord that corrupted my (local) git repo and almost made me lose a day of work. Even with source control there are risks.
If there is anything positive to take away from this,it should be to learn to consider risk management.
If you write code even, say, once every week and haven't yet bothered to learn how to use source control, you owe it to your mental sanity to do it now.
