*** June 15th: Note there has been an UPDATE below ***
Once upon a time, a recluse philosopher, and a Senior Engineer at Google Search, got into an ethical debate on the Twitter, over Google’s policies, implicit bias, and ostensibly their dominance over our economic and civil discourse. To be fair, I will print said debate mostly in full – it is not that long, take a quick read:
So, being the obedient servant of truth that I am, I set to writing the promised “writeup.”
In so doing, I considered what was asked of me again:
Please, point at places where the guidelines encourage what you think is the wrong thing.
I was then struck by the irony. Go through the entire guideline and point out the places where I think they are encouraging the wrong thing?
At first that struck me as an example of Google’s inherent superiority complex. Only the over privileged could be so bold as to demand any complainers format said complaints to their satisfaction.
To be fair, Haahr was not likely meaning to be arrogant (typically the privileged never do). More likely he was simply thinking practically, like an engineer, in that once he has my critiques mired to a specific section of their ethos, it would be easier to take said complaints to the proper departments for discussion.
And in that honest practicality, Haahr and Google are to be complimented: at least he was willing to hear me out! That is one way they likely justify to themselves their operations they commit each day: they think their occasionally willingness to listen, at the occasional search conference, or on Twitter, makes them the good guys… Or at least, good enough…
After all their motto is (or used to be): Don’t be Evil. Now, do the “right” thing…
Well, I am here to say (once again): no. They don’t. They are a form of evil. A very benign evil, to be sure, all evils weighed equally. But evil none the less.
And even the most benign evil, can tally up to major consequences.
The following is an invective against the implicit bias located in the Google Quality Rater’s Guidelines and a list of practical recommendations to fix it. The guidelines are very long, and my need to eat steep (as is all of ours). Thus, the amount of time I can take away from productive hours to do philosophy, is sadly, and unethically, short. Perhaps like Socrates, I should argue Google hire and house me full-time, to do nothing but question and critique them.
This, of course, is what Socrates requested his punishment be, right before they put him to death…
The quality raters guidelines essentially is, of course, just that: a guideline for their human raters of Google’s search results, to try to guide their ever-developing AI, to ensure said results remain of the highest “quality.” They are, as he says, quite long. I have read them, when they were leaked privately to me, long before they were made public.
Highest quality is, of course, merely a euphemism.
And at the end of the day, they are just rating websites. Our websites, to be precise. As Google search results consist of nothing more than our content they scrape and “borrow” and put into their database, package, and sell. We are Google’s product. From the websites they package and produce for searches, to the information they “lift” from you from your searching and browsing habits.
In short, they sell us, to us.
A digital pimp, they slut us out to marketers, who in turn sell us our secret dreams.
The problem is, in Haahr requesting I tell him what is wrong with his rater’s guidelines, in effect, he is just asking me “How do we rate websites better?” … in fact… “How do we rate businesses better?”, thus side-stepping any questions, like, whether they should be the ones rating businesses in the first place. Whether they should be the ones selecting what businesses are “highest quality” and thus get traffic in the public search utility. And resultantly, whether they should be the ones destroying businesses too, without any oversight. Without any regulations.
Businesses, I will remind you, that are tied to real live people, who have jobs (or who had jobs before Google removed their website from page 1, and they lost any appreciable internet search traffic, and thus their business), and were trying to participate in a healthy, and what they thought was open, economy.
The digital economy is not open. Any of it that passes through Google, has been privatized.
Google hides the exact stats But from what has been admitted by them, Google serves anywhere from 40,000 to 63,000+ searches per second.
And Haahr’s jump, him challenging me to go straight to merely showing him where in his rater’s guidelines they are not actually policing for high quality websites, that’s the problem.
As I told him, I do not necessarily have any issue with the how they define what makes a website “high quality” or not. At the end of the day it is all just user clicks anyway. If the users of Google seem to like what they are presented with, that website (and thus business) gets promoted, ultimately. Google watches what everyone clicks, everywhere they can see (which is far more places than you would expect), and makes their adjustments thereby.
Of course, because of this, Google likes to argue they are a fair, democratic search system.
But sadly, this is a lie.
Google is not a democratic system.
A democratic system has an official vote, for official, explicitly stated reasons, for official, explicitly stated, outcomes.
You know what you are voting on. And that there is a vote. And the reason why the process of voting to govern ourselves is important.
This is not what Google does!
Google is simply the digital rule of the callous mob. Lazy indifferent hoi ploi, just wanting their fast-food fix for information. All spied on and then packaged up by a private, for-profit, elitist unrepentantly American, Silicon Valley, delusional big-tech, “don’t be evil” money grubbing, capitalistic, corporate lie.
They maintain they are benevolent dictatorship. These are the rules we have chosen.“It’s the decision we’ve made.”
However, as I tried to tell him above: the problem with even the benevolent dictator (whose decisions you may agree with completely), is that the system is still dictatorial.
Dictatorship as a system to govern ourselves has been (or is supposed to be) abolished in the West for a reason.
Because it is insufficient. Because it is unwise. Because it is unethical, as it is dangerous. Eventually hurtful.
As it turns out, the truth is far more delicate, and far more important, than leaving it to the hands of corporate capitalism, however idealistic and convincing that they “Do no evil”.
Their very existence, their very position as truth arbiter, is evil!
There is a reason Plato ultimately argued against the philosopher king.
Because truth is simply too delicate to be dictated by one entity.
That is too much power, for one, single, public servant.
No matter how good they are.
And they are never that good, are they? They make mistakes.
The only mistakes which are tolerable are the ones a democracy makes. In a fair, official, vote.
Because we voted.
And, as it turns out, that is not only the best way to govern us.
But to govern social truth: who gets to say what… Who gets to hear what.
Now, the democracy of the West has more or less collectively decided on the former, (who can say what) to be mostly anything, and everyone. That is, if you can afford a computer and twitter…The very poor never get to say much of anything, of course. With all the freedom of speech in the world, the rich who made up said rule knew all along it was irrelevant, because it also never guaranteed the capability.
But we have foolishly decided the latter (who gets to hear what) should be secretly determined by a private (actually, a cabal of private), for-profit, American, Silicon Valley, corporations.
Without any oversight, what-so-ever.
Without any regulations, as to who/what they censor, according to their private, personal, values.
Without any guidelines for their guidelines.
That’s what you need Paul. You don’t need me to fix your guidelines.
You need me to provide guidelines for your guidelines…
Where exactly do your guidelines go wrong?
Your guidelines are insufficient in capturing The Good.
Because you don’t even know what that means.
Because you don’t, or shouldn’t, have the privilege of privately (and surreptitiously) dictating it to us.
True, it is only discovered by the unfortunate, lonely, human who deletes all other opinions, and seeks it with soul crushing honesty.
And it is only implemented by a liberal democracy, guided towards it, but never over architected to define it as “such and such.”
Because the Dao that can be mentioned, is not the eternal Dao.
The particle that can be pinned down, thus loses it spin.
Because we murder to dissect.
Because you cannot measure infinity..
So, that leaves us with the only justifiable practical implementation of The Good, a. K. A., the golden rule of liberal democracy:
You are free to live your good life, in so far as you leave everyone else unmolested to seek theirs.
…For who the fuck are you, to define what my good life is for me?
Therefore, this is what Google needs to be: be more liberal democratic.
Liberal Democratic Regulations for the regulations.
What does that mean?
That means: Referendums.
An official vote. On official things.
A charter of rights, we help determine, for searchers.
A charter of rights, we help determine, for the sites/businesses/information you “borrow” to sell back to us.
A charter of rights dictating what, who, and when you can censor. And a more transparent way to see and control on the fly what, who and when has been censored.
A more transparent, assured way to get out of said censorship.
To be innocent until proven guilty of your “manual actions”, your “algorithmic demotions”, your core ranking “adjustments”, a. K. A., when you personally and secretly decide what businesses get to have the traffic flow of the majority of the internet. And which ones do not. Like a power company secretly deciding who gets to have heat in the winter, and who does not, without any government oversight. Without any warning. Without any chance of representation. Without any chance of appeal.
A chance for us to officially determine what we consider to be a “low quality” website and thus, ostensibly, business.
Alternate listing mechanisms: perhaps alphabetical? Perhaps random set of worthy enough businesses as determined by rules setup by us, the public you serve? Not always just the big brands? Maybe you could have mom-and-pop Tuesdays?
Why are all the big brands getting all the traffic on the big terms anyways? How are we to know Google does not have a private relationship with these other big brands to promote them in search?
The only way to know this is liberal democratic controls.
A more official and deeper say into what we consider to be the “right” answer, for whom.
Control of our search database, not yours. As it is nothing but a copy of our websites anyways.
And finally, a public representative, a public advocate, working at and for and paid by Google, voted in by the public (not google, the public – your user base) for a yearly term, to speak for the public, and to fight for their needs from Google’s privately controlled public search utility.
To help you, by questioning you.
What Socrates demanded his punishment be.
Before they killed him.
This, my dear Paul, is how I would fix your guidelines.
The problem is not in your guidelines.
It is that they are yours.
And not ours.
Romeo, Oh Romeo, What’s in a Name?
The problem all along was not in your guidelines, but in your guideline. Your original guideline.
Don’t Be Evil.
That was ill said.(and you did not fix it by then changing it to “do the right thing” this made it worse, far worse).
Your guideline is, and should have been all along:
Don’t be Hurtful
I repeat: don’t be hurtful…
This, this, was what you were searching for.
This, this, is what would guide you.
This, this, is what justifies and excuses you.
This, this, is The Good: Provide as much value as you can without hurting anyone or anything.
Go that far, no farther.
Where there is hurt required (like in demoting sites / censoring / hurting businesses, which yes if ranked, would be necessary) the society so affected must democratically decide a bill, or charter, to follow in that case. A charter that is proposed by, perhaps wise humans who know and love The Good for everyone, and then voted on and finalized and adhered to for a year, (at which times it comes up for possible adjustment).
Then it is not you (Google) hurting anyone anymore (in order to profit from our hurt).
It is the body politick…
What we all agreed to.
The only body allowed to do such a thing.
The only justifiable way to do it.
Why? Why would we do such a thing you silly philosopher?
The reasons why Google might spontaneously decide to do any of these things I suggest is this:
Regulate yourself according to liberal democracy, before liberal democracy regulates YOU.
You have already seen Facebook called to the carpet (literally).
Your time is coming. The techo-idiots are dying, and people who understand who you are and what you do, are coming into office.
I am doing my best to inform them. To warn them.
I would recommend, to maintain your decaying good-guy image, and somewhat magical, cryptic, branding, you spontaneously choose to police thyself in a much less capitalistic way, and much more liberal democratic, forgiving, truly good, way.
Because it is only a matter of time before we do it for you.
Paul, I want to make it clear, from my online discussions with you, and my friends in the SEO industry who have met you and talked with you in person, you seem to be a straight shooter.
So please do not mistake my passion here for anger, vitriol, or cruel irony.
At the end of the day I am a tiny little man, trying to argue with Leviathan.
Trying desperately (and, I suspect, futilely) to convince and persuade that maybe Leviathan is off course. And always has been.
Maybe there is a better way to organize yourselves.
Maybe, I should not be pointing out where the guidelines encourage the wrong things…
Maybe the society you serve, and have the capacity to harm, should have an official capacity to do so?
Show the world how good you are. Show them how, when Facebook only paid lip-service to the criticism of “the city”, Google instead decided to regulate itself, and implement massive policies and mechanisms where their entire user base could have an official say in the running of the search engine, at different levels in different ways. An advocate voted on by the people, paid for by Google, who does nothing but represent said people. A charter of search rights for the searcher and searchee, that can be modified and fixed every year to some rational degree. Official votes proving you care about your user base. Google advocates, like lawyers, who help you in any manual action cases against Google. The right to control our data. And the other suggestions I recommended above.
If you regulate yourself, you will be a future thinking beacon of what big-tech promises it is. Or is supposed to be.
Enlightened. Altruistic. Better.
If you don’t, you are nothing but the money grubbing, greedy, capitalistic, American hegemonic, hypocrites – a mockery of true friendliness, using our secrets to profit off of, and bending the truth ever increasingly to your will – that the rest of the world is starting to realize you are.
So, if you don’t want to listen to me? Fine.
I’m just some guy anyways.
Listen to your own guidelines, Paul.
Do the right thing.
Or we will do it for you.
It is only a matter of time…
Let’s moot all the ethical pre-debates then. Below I seek to give you more blatant, practical, suggestions.
To be clear, I will no more concede my ethical position than you will yours.
But I did seek to give you practical solutions. This is the challenge. A challenge I intend to meet below.
You do not have to understand my position to understand the practical suggestions, and yes, as I intend to show below, they are quite and very easily implementable by you.
Aren’t they? Your programming team is good enough to implement what I suggest below…?
Are you not?
Fair warning: I still must embed the short-form ethical arguments below merely for other readers who will not go back and read my initial argument. Please ignore them, Paul, they are not for you.
They are for history.
They are there merely to elucidate and guide the practical suggestions, which are for you.
And so, to that end:
You are asking me “What are wrong with my guidelines?” this is a stage 2 question, when you have stage 1 practical problems that affect the stage 2 practical problems.
Thus the stage 1 practical problems must be dealt with first. After stage 1 is dealt with, then I can tell you what is wrong with stage 2.
The stage 1 practical problems are as such:
In short, your company, and thus search engine, is evil. Evil in structure. A dictatorship. Somewhat benevolent, yes. But a dictatorship none the less.
A dictatorship is insufficient. A dictatorship is in and of itself is evil. Evil in structure. Risky. Hurtful. Fallible. Dictatorish.
It needs to be more liberal democratic, if it is to be tolerable. We do not accept dictatorships in our political organizations, and there is no reason a priori we must, or should, or ought accept them in any corporate organizations either.
Especially ones with the overarching influence and public service you provide.
Sorry, you have an essential public service (a. K. A., a public utility), and you censor, whether you wish to admit it or not.
Every honest person can admit this. You simply don’t want to for personal and professional reasons.
But fine, I don’t care about that, the suggestions can be implemented whether we agree with the ethics.
And so, the rub is this:
Only a Liberal democracy (LD) through a periodic, announced and scheduled (on your website main page), yearly, referendum (REF) process, has the ethical authority to decide what gets removed from / censored / hurt by / search traffic. Yes, the website is yours.
But the searchers are not, sir.
The search traffic is a public resource. Only the body politick gets to justifiably wield it. Direct it. Control it. Censor it. For good or ill.
Further, the website is yours, sir, but the data is not. It is ours. You “borrowed” it.
The LD through a REF process needs to control what you take; when you take it; what you give; when you give it.
The referendum is advertised and scheduled. Once open, users have 24 hours to vote on these matters. They answers questions like in your rater’s guidelines.
However, they are also warned this is an official referendum. And that their answers will help shape search rights for searchers and searchees (i. E., who will get censored, and why, and how badly) for the next year.
IMPORTANT: They are also explicitly warned that these choices may have consequences for 1) some kinds of information, and 2) some kinds of businesses. They make their choices from multiple choice. These questions would have specific questions, just like in your guidelines, like this: should we be more harsh, equally harsh, or less harsh, on this kind of page?
(these ARE specific practical suggestions you could implement in your guidelines now: it fails to advise the rater of the ethical and social risk of being too punitive)
And meta questions, like this: Do you want mom and pops and smaller businesses to have more search representation, the same, or less?
Do you want to see dirt on people more, the same, or less (remembering that this might someday include YOU or a loved one at any point if something unfortunate happens, like a DUI, etc.)?
All the relevant questions for ethical ramifications are asked. The LD decides.
Thus, the charter is set, and stands for a year. It guides your decisions either philosophically (in your internal engineering meetings), or direct automatic tweaks of an AI you build, to adjust your rankings – sorry correction – to adjust the LD’s rankings, to the LD’s effect, as the LD has demanded.
You provide the search? Fine.
These are our search preferences.
You already ask us for our personal search preferences.
Well you never formally asked the LD for theirs. This is the form such an answer must take.
Of course, if the LD is harsher in their voting, then the rankings are harsher. Google is not to blame, theoretically, you just implement what charter the LD demands.
But also please notice this, it is called the LD, and not the mob, because the guidelines never can be that harmful. The guidelines are not just democratic, they are liberal democratic.
Searchers and Searchees in the system have rights!
(That is what your guidelines are missing. And a mechanism for the LD to generate/adjust them.)
In short, in being liberal, the charter is predicated on The Good.
You keep bragging you don’t know what the Good is. That is a bug. Not a feature.
This is to your shame.
I will tell. Again. It is this: Harm no one, optimize value.
Mitigate risk with all diligence
Seek what’s idyllic, with all care
The libertarians are wrong: freedom to an ill end, is not good...
This is justified liberty: freedom to define and seek your Good unmolested.
The only sanctions and hurts we allow amongst ourselves are the ones we consciously voted on.
Because at the end of the day, each of our good lives is all equally important to ourselves. So they are all equal in vote.
Thus, to be justified, the LD charter contains the Good in LD as a basic philosophy (making it the LD charter, and not simply a mob charter). The LD is protected from its more mercurial moments in history, and any yearly changes are slight. The LD is always reminded, that they can run afoul of their own choices. Unlike Rawl’s “Veil of Ignorance” we implement the opposite “A Veil of Reason” – Like Locke’s golden rule of justice: the reminder that these decisions bear consequences, that this can hurt you too. Or your loved ones. So don’t be so quick to be harsh on searchers or searchees…
For example, as a result, they might be asked second order questions like: WARNING: if you make these search choices, big brands will get the majority of the search space, and mom and pops will be statistically demoted… this is will cause less mom and pops to stay in business… is this what you think is fair? Is this what you think is right?
WARNING: if you make these selections, you have voted to prioritize scientific truth as the primary verification method search will use. This might tend to exclude, or even offend, other forms of faith? Do you think this is fair? Do you think this is “the right answer”?
** However much I personally agree that scientific truth is truthier, and that religious dogmatists needs to go crawl in a hole and stop ruining history, it is not my nor Google’s authority to impose our views upon them, any more than it is for the religious and dogmatic to impose their views upon us.**
So, the REF process questions are worded as such that it will naturally incept the correct amount of empathy/goodness/fairness/reasonability in the voters (whatever that is – I am not dictating. And neither can you.. see? do you see the difference? do you see the long arc of ethics here? Do you see how far yet you need to go? As Rousseau once said,do not blame me if being ethical is hard. This is the way it is.).
Thus more balanced, empathetic choices, are likely, mitigated against our desire to not be bothered with what we consider “spam.” Please note, it is what the LD officially considers “spam” (when they are officially asked, and reminded of the consequences) that is important, and what you are ethically justified to fight. No more, and no less.
Not what you (as a private, for-profit, biased, unrepentant, American corporation) judge as “spam.”
We are not all American. We are not your selected biased group of Internet Elitist raters either. We do not all share your views.
(Do you really want your answer to be: don’t like our ‘Murrican search engine, build yer own? When we have Russia and China doing just that, and censoring far worse than you? Do you really want the competition? Facebook has already been called to the carpet. Preempt where the Zeitgeist is going. Implement your Don’t be Evil mandate, save us from ourselves, and implement the more generally LD search engine for the LD West, and be the shining beacon of what big tech is supposed to be, or go out of business and be forgotten in 50 years).
And so, to reiterate, the mob is not the LD. We cannot accept even what your private raters judge as spam either. For (I assume) they are not being explicitly and procedurally reminded of the consequences of their actions – that they hurt some people in their decisions.
Nor have they been chosen by the LD as appropriate representatives for a term (a flawed mechanism which I do not suggest, as a representative democracy is, at least in this arena, and maybe overall, insufficient and unnecessary given our technical capability).
And naturally, the simple clicks of the mob are not sufficient to ethically justify what “spam” is either, because nor have they been reminded of the consequences of their clicks.
PLEASE NOTE: Votes must be official, announced, and deliberate – the pros and cons spelled out beforehand – if they are to be used as ethical justification for any guiding of any system.
The vast majority of us do not even realize you are spying on us, and making decisions with ramifications for them and their fellow citizens, thereby.
** Whenever people know they are being watched, they act more ethically, typically.**
This is, again, a feature, not a bug.
And so – please inform your PR department – to be used as a valid justificatory principle, a voting mechanism must be official, reasoned, pros and cons weighed and spelled out, and deliberate. Everyone has to know what they are voting on, and what’s at stake.
Then their vote stands for the year. If we don’t like what LD voted for this year, we only have to wait a year, and another digital referendum on our search rights is held, in which we can tweak up or down the harshness of your search algorithms.
Right now your raters are like the UN, making decisions for us in our country, that we have no mechanism to change.
Also what is required is a more transparent, assured way to get out of your censorship box.
I suggest you adopt something closer to the LD legal process for manual actions.
That websites are considered innocent until proven guilty. That they get a chance to defend themselves or fix things BEFORE the manual action is applied.
Why be punitive dicks?
Seriously. Simply warn them if they do not comply in 30 days a manual action will be applied (and then watch what they do and make your anti-SEO algos accordingly).
Why not warn them in search console that they will likely be affected by an upcoming core adjustment and to make their website better.
At least they can plan for their businesses’ destruction?
Wouldn’t this be the “right” thing to do?
No, you don’t have to tell them what to fix exactly, and thus give anything away, as if that was a valid argument anyways. (It’s not).
Still, even warning them without telling them what exactly to fix, is still better than springing it on them.
And you could, in that search console email, point them to the TC forum where they will get some suggestions at least, that might actually cover what the new adjustment might be looking for.
I would even go so far as they get the right of representation in your system, a google advocate there to defend them and help them get their site out of violation of the charter.
If not also what I said I my original post, a “Socrates”, on your pay, to do nothing but criticize you, and keep you on the straight and narrow.
Can you imagine that? Can you imagine having an employee on the payroll, that answers to know one, and whose sole job is to question and critique you? Can you imagine how ethical, how good, how forward thinking, how transparent such a company would look?...
Fairness in Implementation
I have covered how to introduce fairness and justification into your guidelines ahead of time.
But what about after? What about during implementation?
Some form of transparency of what important sectors or demographics (by class, race, structure, political leanings, and whatever other identifiable demographics) of websites are getting search benefit. And which are not. Which are being demoted? Which are being censored?
No specific site needs to be mentioned.
But a set of statistics something like:
Big brands: 89%
Small business: 8%
*margin of 3%
Etc., would be very instructive in making ethical voting decisions when we go to adjust the charter of search rights every year.
To be truly fair, of course, a third, NPO party should be given access to your raw search data so they may confirm your statistics. To keep you honest. They can take the task on of publishing this.
So we know you are in good faith, not betraying our data and trust in wielding so efficacious a power in our LD, but instead actually implementing our official search preferences.
Thus securing your moral high ground.
I know, the way I feel you talk down to me on twitter – the way you simply deny certain facts, the way you demand I form my critiques to your satisfaction, the way Google Fan-boys jump in liking, commenting, attacking me – you think you have the moral high ground.
But that’s just the American in you talking.
To remain ethically justified, with such a public service that you wield, would require something like the steps here.
To quote the great Stan Lee, “With great power, comes great responsibility.”
Due to your concordant power, this is what form your responsibility must take, that is, to remain tolerable to the LD.
If Men Were Angels
In your response to my initial argument on Twitter you quoted Madison, one of your founding fathers, “If men were angels...[no government would be necessary.]” You used this to argue that, because black hats are not angels, that rules and protections against them are necessary. Fine.
But if you had only finished the entire quote!
And there is the rub. With all the hurt Google causes, Paul, and all the hurt Google could cause in future elections and to future businesses, you are no angels.
Thus, by your own reasoning, external and internal controls are necessary. On You, too. On Google. From us.
It goes both ways.
And what those rules against Google must be, to remain ethically justified, I have submitted.
I repeat: You cannot have it both ways, Paul.
You cannot say to me, men are no angels, and thus we need to enact controls against them in our search system.
But then deny the exact same controls, the searchers and searchees, need to have over you.
And so, to say you are not a brilliant enough programmer to understand the practical changes I have outlined here, or implement them, is simply a convenient lie.
We both know what I have suggested here for your programming team, would be quite easy to accomplish in some form or another.
Please don’t insult us by pretending otherwise.
Thus, the only thing that remains is your will to do so.
How good you are.
How liberal democratic you truly are.
So, let’s see how good you truly are.
You’ll forgive me if I don’t hold my breath.