Slashdot Mirror
Genetic Access Control Code Uses 23andMe DNA Data For Internet Racism
rjmarvin writes: A GitHub project is using the 23andMe API for genetic decoding to act as a way to bar users from entering websites based on their genetic data — race and ancestry. "Stumbling around GitHub, I came across this bit of code: Genetic Access Control. Now, budding young racist coders can check out your 23andMe page before they allow you into their website! Seriously, this code uses the 23andMe API to pull genetic info, then runs access control on the user based on the results. Just why you decide not to let someone into your site is up to you, but it can be based on any aspect of the 23andMe API. This is literally the code to automate racism."
1: Stop caring what other people do.
2: Don't have a 23andMe profile (accessible to random websites on the internet, or at all) if you really care and think this will actually affect you.
Perhaps some users will implement it in a harmless and beneficial way, such as creating a safe space for women. But it’s just as likely that, in a few years, Googling for a snippet of this code yields search results that are the equivalent of a who’s who of racist and misogynist sites.
3: Take your sexist, racist agenda and go the fuck away.
Now you won't just have to buy other people's pee to pass your drug test, you'll have to buy other people's spit to get access to racist websites.
Finally a way to avoid offending certain people and gorillas.
I'm an American. I love this country and the freedoms that we used to have.
So visitors to his website:
* Must have been sequenced by 23andMe
* And be so interested in his website that they are willing to give him access to their genetic data
* And meet whatever genetic filter he has imposed.
At this point, what he is running is less of a 'website', more of a 'diary', as it will have only one reader.
Quattuor res in hoc mundo sanctae sunt: libri, liberi, libertas et liberalitas.
It is easy to see how such code would be used to discriminate based on issues such as race or ancestry. With that in mind imagine how many skinheads, neo-nazis, klanners, and other groups will use this to restrict access to their website and their members will have to get these reports in order to have access and then discover that their DNA isn't as pure as they first thought.
I didn't realize that "race" could now be determined by a genetic sequence (or two). If true this may lead to some very troubling possibilities.
It would be possible using CRISPR (a recently developed means of precisely targeting an exact genetic sequence) to have a virus that could infect just one particular population. Smallpox comes to mind because 1) very few people in the world are currently vaccinated against it (it was made extinct a while ago) and 2) its DNA has been sequenced and published online. So, using a DNA synthesis machine you could make a version of this virus that would target a particular population.
I believe such machines can now make DNA sequences long enough to create viruses. I also remember someone creating a much more lethal version of smallpox that could kill all of the laboratory animals it infected, including ones vaccinated against "normal" smallpox (I think it was 100% lethal).
Of course, making a virus that would go after a particular Sex as opposed to a racial characteristic should be much easier (just target any gene found on one of the sex chromosomes and not the other.) This particular scenario was explored in Frank Herbert's book "The White Plague". A related scenario might be the film "Children of Men" where all the women(?) are made sterile.
..and I suppose you know the identity of this supposedly objective arbiter of social justice who will decide how technology will be used?
Your shitty argument is little different than the "this is why we can't have nice things" retort often repeated here.
I think the author, using the "offensive-computing" nick, knew very well that this would trigger a discussion and that's probably the reason this project was created in the first place.
For this to work, there are two required components: code and data. The code has already been created and if it hadn't been, implementing OAUTH and using a REST API is within the scope of many developers' skills. So the route to avoiding abuse of this technology is by restricting access to the data. Simply put, don't give any web site access to your genetic data.
Now only people with a 30% chance of being able to smell asparagus in urine will be able to visit my site!
Surely you mean "orange haired orangutans".
This has got to be against the Terms of Service, right?
The problem is that with enough uptake it's no longer a self-extinguishing problem, but a shutting-you-out problem. No DNA profile with that third party website? No access to facebook, google+, whatnot else. With even governments (like, infamously, the UK gov't) considering using facebook profiles to "authenticate" citizens for access to government services, it's easy to see that this may well end badly.
And you can't really leave this to "the market" because it inherently shuts out the negative control feedback, since shut-out people have no voice. A bit like the breakage and shutting out of visitors by "upgrading" your website to cloudflare -- and cloudflare deciding some visitors are really spambots even though they're merely behind CGNAT that cloudflare didn't know about, for example -- that may or may not even be visible in site stats.
Would you mind being an overlooked minority statistic if that means complete loss of access to, say, all government services?
So while "don't play ball" is a good start, it may well not be enough.
What if someone wanted a forum for say, Huntington's disease (just to pick a random genetic problem). What if they wanted a section where only people who truly have Huntington's can share their experiences without getting trolled by just any person? (duh, people with Huntington's could troll too, but you get what I mean). Just to say, not everything having to do with someone's genetic code is "racist". Almost makes me think the OP is just trolling too.
I used to have a good sig...
how secure could that actually be as an access control?
I mean, the access control isn't checking your DNA. It is checking if you have access to a genetic profile of someone with matching criteria. Or, as a completly different attack vector: access to body fluids of someone matching the allowed filters. ("whitelisting" gets a whole new meaning here....)
bickerdyke
I thought the narrative was that race doesn't exist? That we're all the same race? How then is is possible that one can determine so called race by examining the genetic profile of a person?
Is it possible we were lied to and that race is indeed something that can be determined as a composition of genes and other genetic data?
The correlation between genes and culture is becoming weaker in the modern world. People have far more interaction than before with people from other backgrounds since travel is easier, many people live in large cities and people communicate online. Therefore, to guard access to an online community based on genetics seems like the wrong solution, since a superficial genetic profile will not predict how well an individual will fit into a community.
I agree, because only White people can be racist. Racism is privilege plus power. We need to stop racists from being able to do damage by programming software. They should be restricted from accessing Github and learning how to write software. We should round up all White programmers and scrutinize their political beliefs so that we can eliminate the poison that is racism from our society. It's the only way the future will be free from racism.
But only white people can be racist. Racism = privilege + power. Only white people have privilege, whereas the People of Color lack privilege and social power. Therefore it's impossible for a Person of Color man to be racist. We must unite against all racists and purge them from our society. It's the only way the future will survive.
You sound like a racist. You should be ashamed, and then you should have your identity tracked down so you can be made an example of as the people of tolerance and peace have you removed from your job and your finances ruined. No racists on our streets!
> A GitHub project is using the 23andMe API for genetic decoding to act as a way to bar users from entering websites based on their genetic data — race and ancestry.
This is very useful. Why should e.g. a syrian webforum allow access to jews, so that they can spew hasbara propaganda and recruit traitors to work for the Mossad in support of the saudi proxy's ISIS invasion?
Why should e.g. a european webforum allow access to russians, so that they can spew militant panslavic putinist propaganda, deny responsibility for MH17's downing and the Katin massacre, while recruiting traitors to work for the FSB/GRU in support of the the upcoming muscovite invasion of the Baltics?
Why should any web forum located outside of Africa grant access to nigerians, so that they can spew type 419 advance fee fraud, while calling whites stupid (even though the average adult negro has an IQ of just 70, which is considered slightly mentally retarded for whites, who have an average IQ of 95-100.)
It is a matter of fact that human races are not equal. Those nations/people/tribes that have no neanderthal cross-breeding heritage, because they never migrated out of Africa or left the black continent too late, haven't received large-brain encoding genes from the Neanderthals and this means low IQ for them. The only people equal to the white caucasian race are the japanese.
(Although one must mention that south koreans and iranians-persians have the potential to grow equals. Especially the ancient culture and philosophical heritage of the persians is underestimated by the western world, which is blinded by jewish propaganda. The Iran embargo doesn't work, exactly because they are a coherent, utmost patriotic, resourceful and clever nation and they do work, unlike arabs. Yet the world knows so little about them, because of the embargo.)
Like this white supremacist who found out that he's 14% African.
I can't think of any practical use for such a Genetic Access Control method nor of a reason to feel outrage and clamor "racism". For a start, this app only works for users who are also 23andme clients anyway, who also agree to have the app access their data (à la Twitter), and I'd say those people pretty much already explicitly waived their genetic privacy.
Also, I can attest to how widely inaccurate some of the results you get through the API can be, especially the ethnic origin results. In my case it's ~16% inaccurate. It's known to overstate european origins and downplay or entirely lack quite a few less common origins, as the comparison database misses data from entire ethnical groups (Sinté, Romani...) or has only a handful (or single) individual DNAs for many potential origins (Romania, Azerbaijan, etc.). And the haplogroups tree that the API reports from is outdated (and, again as in my case, very lacking of resolution in several branches).
Maybe we deserve this world ?
IF haveOnlineGeneticProvile
THEN PRINT "You are too dumb to be allowed on the internet."
I do not fail; I succeed at finding out what does not work.
A sensible government would create its own authentication infrastructure instead of relying on an external platform. Unfortunately, the UK government hasn't displayed a lot of sense lately when it comes to privacy.
As someone without a Facebook account, I know it can be problematic when people and other companies just assume you'll have no problem giving up privacy for a little convenience. I think though that a genetic profile is considered more private to many people than their chats and photos, so they'll be less likely to hand it over just as easily. For example the sharing of electronic medical data is something that generated a lot of resistance here in the Netherlands.
In any case, trying to stop this at the technology level is a lost cause, since it is just too easy to write the code. If you'd want to stop it with regulation, you'd have to do it at the policy level: forbid sites to use it. But looking at ineffective half-measures like the EU cookie directive, I have more confidence in end users rejecting genetic access control than in legislation blocking it.
Disagree. Any speech or expression should be allowed, or your remarks
themselves could be considered hate-speech. The Constitution's framers
understood this fine point that speech that is not acted upon is nothing more
than speech. Now if a person acts on their words, then yes, that's a problem
I thought the narrative was that race doesn't exist? That we're all the same race?
Yup. We're all the same specie. "Race" is just an arbitrary label used to make convenient distinction among a group of individual that would other wise mate together : e.g.: breeds of dogs, etc.
How then is is possible that one can determine so called race by examining the genetic profile of a person?
Is it possible we were lied to and that race is indeed something that can be determined as a composition of genes and other genetic data?
Short answer: Nope, it's not possible.
Not so short answer: Well depend on what you define as "races" i.e.: you'll need to constantly redefine your arbitrary "race" label along as the set of gene that you try to use to map it. With that definition of "race" drifting farther and farther away from what biggots with an agenda would like it to be.
Long answer: Remember that the only valid real barrier is *specie*. i.e.: between groups that *cannot* mate together. Inside that division, all individual can mate together. Humans are specially good at that "mating" (well, maybe except some basement dweller): because from our dawn we've been one of the most mobile specie, only bested by birds and some big aquatic animals. Humans do tend to get around a lot. As a result our specie is constantly under heavy genetic mixing.
You can put some arbitrary labels (hey, let's put together individual that have darker skin tone in one big arbitrary bunch and call them niggers!)
You can even find some labels that are actually convenient (statistically, people who typically come from Caucasian ancestry tend to have a bit higher probability for some disease and a bit lower for others).
But if you got into the details, it starts to get muddy, you'll find that some people that you put in one of your arbitrary categories come from completely different lineage that split quite some time before, whereas other that you wanted to put out of that label actually share much more recent ancestry.
(That white supremacist biggot over there ? He want to be put in a special category, aside "non whites" ? well on the surface it might look like some vague idea like correlating it with skin pigmentation. Except that among those non-white he'll probably put native american, but also african and north african. And his grand-dad happens to be from an italian immigrant family, and over the course of history southern european have had a lot of exchange with northern africa. See where I'm heading ?)
So, in short what works well to separate breeds of dogs (which are bred in very controlled manners and you can somewhat keep something like a breed more or less pure) absolutely doesn't work with human that fuck around a lot.
Same for the genetics.
You'll be able to spot a few marker (let's say: a gene that helps control the base amount of melanin in the skin). But once you start using this, you'll notice that a lot of the "wrong people" end in the "wrong category".
That specific list of markers you've assembled together, will consider some of the white supremacists as non white (because of the complex mixed ancestry most people have) while failing to cast aside one of their usual target.
An idealist (like me) would dream that this would help some of the biggots to realise that we aren't that much different under the skin, we're so much mixed that there's a high probability that the biggot has a bit in their blood what their usual target is.
Saddly, the reality will be that they'll be still endlessly tweaking around their genetic definition of "race", to constantly keep itt more or less matching their agenda. They'll never have a definite set of markers, they'll constantly need to patch them. And over time it will evolve to something that at the genetic level doesn't remotely look like what they pretend to be (by the time they manage to include all their friend and exclude all their targets, melanin-related gene
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
The only people bothered by this are the silly gits who have championed real names on the internet, the people who ignorantly believe they are only allowed to have one identity on the internet, and those foolish individuals who broadcast lives to the world.
Online IDs are cheap. They don't cost anything and you don't need to find a shady character in a seedy print shop to score one.
I don't understand what the poster is talking about. According to 23andMe information about access from their public APIs, "Each app will request that you log in to your 23andMe account and provide access. Each app will specify the data it is requesting and what use it will make of that data. If you don’t want to give a particular app access to your 23andMe data, simply deny access to that app." So, a website cannot just go and check any genetic data. Sounds like nonsense to me.
Who cares how this is used? You can use an image hosting site to take a picture of yourself with your username much more easily and be vetted by administrators if they're so racist or sexist that they care about being "pure". I can't possibly imagine that a racist or sexist forum would be so swamped with new users that it would be worth it to automate the process at the cost of turning anyone away.
1) Kuwait has introduced a law mandating DNA testing for everyone, already. https://www.hrw.org/news/2015/... - it's required in order to get a government ID, and a government ID is required to live, basically. 2) In America, at least, the real problem with this code (and likely the real use case) is not blocking access to a website as an expression of "racism" (whatever that even means in this context), it would be using that profile to serve up content _selectively_. It's already been shown that Google gives different results to searches that include "black" names vs "white" ones: http://thevisualcommunicationg... - they could with perfect ease include this DNA data as an input signal to ad selection also. Searching for auto insurance? Your color blindness will cause the search to show worse deals. It's axiomatic that you should never, ever make unchangeable information with abuse potential of this sort accessible to anyone, if it can possibly be prevented.
What on Earth does the NY Post think "secret" means? Not only does the administration admit they're making this database, they've said it will be partially open to the public. That isn't a secret database. That's the opposite of a secret database.
I find your comment offensive and a micro-agression against me and my kind. This is hate speech. Please ban and delete immediately.
We get it - you are scared and lashing out like a petulant child at anyone or anything which threatens to shed just a little more light on your insecurities.
I think the author, using the "offensive-computing" nick, knew very well that this would trigger a discussion and that's probably the reason this project was created in the first place.
Yes, mod this up as the most insightful contribution to the thread so far. Although the author is careful to use other example in the readme, the code and example application are (provocatively) written to discriminate against everyone except the usual racist definition of 'white Europeans'. The 'European' group of reference populations defined by 23andme would normally include Ashkenazi Jews:
https://customercare.23andme.c...
but this group is explicitly excluded by offensive-computing:
https://github.com/offapi/rbac...
Apparently someone forgot that it was a movie, not an instruction manual.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
Because fuck those rednecks, that's why.
Ironic isn't it. That simple statement there is the biggest part of the problem.
Why can't people just accept that other people have different beliefs and stay away?
Its one thing when a racist fuck forces something on you, its an entirely THE SAME THING when you want to force yourself on some racist fuck.
And just for reference, it is not illegal to run your business in a racist manner, like 'whites only', its just REALLY difficult.
Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
NBA, Comcast, Toyota, Obama, etc. care though. :P
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
Depends on the speech, the person, and in what role their speech originates from.
For ex:
-money in politics: not free speech
-private citizen being racist: free speech
-government official being racist in his official capacity: not free speech
The guy who said the election was rigged won the presidency with the second-most votes.
Just though I should point out that the study you quoted is invalid based on the following.
They used names that were already associated with a particular race. How is something raciest if the race its self is choosing it. So I pick a "black" name and search for it, the problem here is that he name is already identified as being associated with black individuals. This is like buying a bag of M&M's and expecting Skittles inside. Yes I know it's a fallacy to make this comparison, but it's the closest example I can think of. This is the same reason we cant rely on any Climate change data, they keep using algorithms that produce the specific results they are looking for. This is a form of data manipulation, not science.
If you want a real example of this, i wrote a simple program that randomly threw Japanese syllables together, then did Google image searches for the results. (this is a very small subset of syllables available in many other languages like my native language English) The results were fairly unbiased, most resulted in nothing useful. But when i did get results I got Japanese name suggestions, Anime, Japanese profiles, etc... Almost no results for Anglo's or Black's, is the search biased. After I was using the Japanese language as a basis, so with logic from the article you quoted Google must be biased against my random generator.
The point is if you are looking for it you will find it, even if it does not exist. Yes there is real racism, but most racism is just cultural differences. For example most interviewers go off gut instincts, well if you were raised in a predominately white environment you are going to have better feelings about whites over blacks, native Americans, or Asians, etc... It does not mean that you are raciest, it's a cultural difference, you know more about what to look for in a culture you are familiar with. This problem gets worse for people who have conditions that place them outside the norm such as those who have ADHD or Asperger's, really you should try imitating those behaviors then try and get hired. Racism is not when you recognize something, nor is it sexism. Racism is when someone knowingly chooses one race over another. It's that choice that makes it racism. Otherwise we are all raciest, sexist, hypocritical, bigots.
Interesting question, though how the "training" occurs is up for debate. It's unlikely to be that, though. It's more likely to be keyed on specific, easily identifiable markers that can be parsed out with deterministic algorithms.
"It's already been shown that Google gives different results to searches that include "black" names vs "white" ones"
And why wouldn't they, and why would you NOT expect to get different results?
Searching for 'Louis Farrakhan' SHOULD give different results than searching for 'Maria Sharipova'. Or 'Lothrop Stoddard'.
Such a vapid argument. Try harder, please.
deleting the extra space after periods so i can stay relevant, yeah.
It's already been shown that Google gives different results to searches that include "black" names vs "white" ones
Wow - You mean putting different words in my Google search... Gives different results (or in the case you linked, different ads, arguably just another type of result)??? Those racist bastards!
Seriously, what the fuck? No kidding, it gives different results! If I search for my name, it gives different results than if I search for Brad Pitt, despite having the same race and gender. Duh.
The world still has real racism. Quit trying so hard to find it in places it can't exist (algorithmic search results), unless you seriously mean to accuse Google of biasing their algorithm to discriminate against people named Shaniqua.
The study, and the related hoopla (this is just one link), was designed to indicate that Google uses a non-obvious, racially skewed signal as an input to search and advertising results. Actually, they probably use more than one such signal. Oversimplified: if I am logged in as "Wilfred Fortescue-Smythe-Smythe III" and search for "bonds" I'll probably see advertising/results for investment vehicles. If I am logged in as "Taneisha Williams" and search for "bonds" I'll more likely see advertising/results for bail bonds. It's important to note that there is NO CHANCE WHATSOEVER that Google designs intentionally racist ad-ranking algorithms; they have too much to lose. What this phenomenon demonstrates is that complex, probably nondeterministic algorithms that sum a buttload of signals, which are designed to exploit demographic/psychographic characteristics to group users by willingness to purchase, can as an epiphenomenon amplify and expose demographic differences in purchasing behavior, some of which differences might cause political trouble with people in a given demographic who compare results with a different demographic. (This works both ways. Some of these "discriminatory" results give higher pricing on things to people in a "rich" demographic).
Okay, people can register DNA samples with 123andme.com and be entered into their genetic database. When you log into a site that uses this API, racist or otherwise, what quick way is there to authenticate? Touch the tip of your tongue to a sensor? This could be a surefire way to identify an individual, so long as the sensor had a threshold test that would prevent people from using the edge of someone's water glass to log in.
pretty obvious you have no damn idea what you are talking about
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof, or abridging the freedom of speech, or of the press, or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances
congratulations your amongst the unwashed that have no idea what the constitution says.
See my other reply on this same branch. That link I gave is just one of several, and possibly not the best reference - just the best I could find in a quick, inadequately caffeinated search this morning. (I was originally reading about airline pricing algorithms when I came across the information I was summarizing there. That led down a rabbit hole that gave the specific searching "bonds" = "investments" if you're white, "bail bonds" if you're black, example).
"Racism" (not the right word for it) CAN exist in algorithmic search results. Please see my other reply in this same branch of the thread.
a way to make sure users of a forum for pre-teen girls are at least female - not men looking to rape children.
The hell did I just read?
Support my political activism on Patreon.
what is up with people abusing the word "deterministic"? it has become a strange euphemism in the computing world, meaning something like "reliable" or "trustworthy".
a deterministic algorithm can fuck up really, really badly (especially when it takes input from untrained or possibly malicious humans), while there are several stochastic algorithms that work quite well, often provably so. even chip cores have layouts optimized by stochastic hill-climbing!
anyway, Google is in the business of showing person A what person B wants, which is usually the product of some person C. unless everyone in the world is non-racist, then of course the algorithms will either "be racist" or suboptimal. it would basically require a strong AI to change this. (A, B, C are not necessarily distinct.)
"They were pure niggers." – Noam Chomsky
no such thing as "fighting words" on the internet. fighting words is speech that can lead to violence or harm in the near future. no violence ever came of arguing on the internet.
also, just because something is fighting words doesn't mean its not allowed. it means that this category of speech can be regulated by the government because the government has a legitimite interest in keeping the peace. govt can regulate fighting words in ways that are directly tied to the government's interest and are the least obtrusive way to achieve that goal.
What do I have to be logged in as to get info on Barry Bonds? Or all the actors who have played James Bond?
The Quirkz Handbook of Self-Improvement for People Who Are Already Pretty Okay
No, it's not. Whether you are offended or not is irrelevant. What makes something hate speech is that it attacks a group, making them more of an acceptable target for future attacks. Hate speech is, in other words, the propaganda campaign that precedes every war and atrocity in human history. That makes hate speech not just speech but also an action, the first blow that starts chipping off the preceived humanity of the target that protects them from violence.
Is someone simply being a jerk? Then it's just speech, offensive or not. Is someone working themselves up into a rage in preparation for an attack? Then it's hate speech.
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
Note that the First Amendment only limited Congress, not the States or other smaller governments, or the courts. One question is that since it amended the Constitution, is Congress still allowed to limit speech with copyright (and other IP) laws?
https://en.wikipedia.org/wiki/Inverted_totalitarianism
No, it's not.
Yes, it is, and everything you wrote following that just proves me right. You are attacking the group of people who fully believe in the concept of free speech.
Now I want to be clear that I do not want you to shut up. But, by your own logic, you believe that your own speech should be censored. And any further insistence that your speech is "ok" will, again by your own logic (not mine, yours), reveal yourself to just be a gigantic hypocrite.
I understand what "hate speech" is. I am merely pointing out that speech about forbidding hate speech is, itself, hate speech against those of us who fully believe in the concept of free speech. There is no way around it. Either you think hate speech should be banned and thus think your own speech should be banned, or you are accepting of all speech. There is no option where you think hate speech should be banned, but your own speech protected. Because your speech is hate speech.
No, you stupid shithead. The search terms are the same. The login names are different.
Reading comprehension fail, much?
FTA: I wondered what a Google Image search would bring up if I typed in "black" names and "white" names.
The author explicitly searched for names strongly associated with a particular race.
But that said, let's play Devil's advocate, and pretend you didn't decide to jump in and start flinging feces without even reading the GP's linked article. If you've logged in, your own login name is a search term. You just don't actually enter it manually. Again - Brad Pitt no doubt gets ads for different products than I do, despite our shared race and gender.
These days, 'racism' is about culture, not "race" in the sense of skin pigmentation or skull shape. We know anyway, that "race", technically speaking, is meaningless in humans.
It could be said that appearance (okay, 'race') is a really crappy proxy for culture, and people choose to discriminate and abuse others on the basis of culture.
By way of example, hardworking, moral, aspirational Hindus and Sikhs are well-liked where I live. Yet, their genetically identical south-Asian radical Muslim kin are despised and shunned, and are dirt-poor as a result. Discriminating against somebody on the basis of a dubious socially-conditioned trait like adherence to conservative Islam (say) would be impossible to do through genetic data alone.
As it were, the worst-behaved and most aggressive people I come across at street-level would be people who wear sleeveless puff jackets, have beards, shave the sides of their heads, drive BMWs, don't bathe, and show no respect to whites or women. Dress and behaviour -- not genetics.
If the site is strict enough it will just be mental incest on that site.
If builders built buildings the way programmers wrote programs, then the first woodpecker would destroy civilization.
Well, I'm a QA guy for a company that makes software used in hospitals. So, to me "knowing the expected output for a given input" is closely correlated with "reliable" and "trustworthy", and so yes, I tend to refer to deterministic algorithms positively. I realize the pressures on an advertisement-selecting algorithm are not the same. However I *ALSO* think that the DNA signals Google (or rather Google's advertisers) would be looking for are the less ambiguous things that are more easily detected with a simple deterministic algorithm. Yes, broader conclusions can be reached by letting a neural let look at a bunch of stuff. But an oncology clinic is going to be more interested in yes/no answers to questions like "does the searcher have DNA markers A,B,C,D,E [which are easily searchable deterministically]", and won't actually ask - or will rather assume the answer to - the underlying statistical question "do markers A,B,C,D,E correlate strongly to a predisposition to conditions that we can charge money to treat"?
Someone intelligent enough to give Google a search term with more than one word.
While Dr. Yak is largely correct on the science, and the fact that a lot of people would be surprised to find certain genetic details about themselves, Dr Yak also completely wrong.
It is completely possible to use genetic details to classify people into groups. There are certainly overlapping people, like Native-American mixed with African, and Northern European. There are lots of people signing up to find out what their "origins" are. There are definite differences between African, Asian, Polynesian, Scandinavian, Jewish, Middle-Eastern, Southern Europeans, and so on. It does not rise to the level of races within the species. Even though: Europeans are largely Homo Sapien/Neanderthal cross-bred, Southern Asiatics are largely Homo Sapien/Denisovian cross-bred, and Africans are largely Homo Sapien/Other ancient hominid cross-bred. There is some mixing of these three, making probably a whole spectrum. Again, though the homogeneity of the human species is so complete we have only one race.
Having only one race doesn't mean there are no scientifically determinate variations. It just means, you can't always tell by looking at someone, what those variations are. I think that is really what Dr. Yak is meaning. These characteristics can be used to include or exclude certain people from some exclusive club. But those doing so, would do well to test themselves first, to make sure they fit in the category they think they do.
If English is, as you claim, your native language then how did you manage to spell 'racist' wrong every single time? It is not like it is an obscure and difficult to spell word... Raciest means the most racy. Racy has not really got much to do with race when used to define differences in culture, genes, etc...
I am not usually one to pick at spelling or grammar, my own is not infallible, but when I see really atrocious and egregious errors it makes me immediately discount anything the person said. That is a failing of my own, I suppose. However, I do not think I can adjust accordingly.
"So long and thanks for all the fish."
neural nets with backpropagation are as deterministic as any other algorithm; if the inputs are the same and no one has "improved" the numerical routines, you should get the same result every time you run the training algorithm. yes, you can add stochasticity to NN training to try to improve convergence, but this is true about many algorithms (whether their implementors realize it or not). running the same trained, frozen neural net on the same input should definitely always give you the same result, unless something weird is happening (as it sometimes does).
i understand that in certain cases, we need provably correct results. my point is that "deterministic" is often (but not always) the wrong word, which i think you've illustrated.
"They were pure niggers." – Noam Chomsky
I don't argue with anything you say, except to note that for the frozen NN you mention, I can't predict what output Y will result from input X, unless I already tested input X. It's not like a deterministic algorithm where output = input + 3 for all input in {-MAX_INT ... MAX_INT-3}. I need an exhaustive map of allowed inputs to expected outputs for the NN case. And product management never gives me one ;)
Damn. I'm in trouble then.
The Quirkz Handbook of Self-Improvement for People Who Are Already Pretty Okay
Are you sure that's not equally driven by geolocation? Or user search history?
Try once searching for 'storage containers ' and see what happens to your results for the next few weeks.
deleting the extra space after periods so i can stay relevant, yeah.
It's hard to tell. You could certainly be right - but, I have to ask, from an "optics" perspective - is that better, worse or meh? To me, it seems silly to entertain the idea of getting up in arms over the output of an algorithm for choosing which ads - that I'm blocking anyway - get shown to different groups of people. But I think the same racism/classism/*ism complaints would be raised in your geolocation scenario by people saying "We've been classified as ghetto residents", right? TL;DR - people self-identify as persecuted victims of large corporations. Which they are, but just not in the ways they're complaining about.
Umm.... everything YOU just wrote proves you to be yet another fucking AC idiot who doesn't have the balls to own up to his free speech.
Fighting Words have long been prohibited. Dressing up your fighting words in fig leaves does not change the underlying intent. Blowing dog whistles still brings the dogs. Implied calls to action ("XXX people are dogs and should be put down like dogs!!!") are still calls to action.
Help! You are oppressing me by not allowing me to attack you!
This comment is my opinion and does not represent an official position of Donald Trump or others I do not work for
The concern trolling is strong with this one...
Look at the bright side. This technology could also be used to produce an even racial balance among the participants at a website.
If a disproportionately large number of people of a racial group have logged in, it will not allow any more of that group until the population is more appropriately diverse.
Alakazam! A more diverse Internet.
There's no time like the present. Well, the past used to be.