Month: July 2016

product management, destroyer of worlds

~ ginsu ~ 3 Comments

God I hate tech hyperbole. And I literally hate the use of “literally” to mean its opposite … but let me explain how bad technology product management can literally lead to the end of the world.

Donald Trump represents an existential threat to humanity. To put such a man at the helm of the world’s most powerful nation is like handing over the controls of a nuclear submarine to a petulant baby. That’s a poor simile only because it’s not an analogy but a nearly literal description.

What can I do about this? I’m just one person, one vote. Moreover, I’m in California, which will surely vote Clinton anyway, so my vote won’t sway the outcome. I could advocate, I could preach to everyone around me, but really most of the people in physical proximity to me already agree with me.

What about technology? I’m in the center of Silicon Valley, I know me some techmology, can’t I do something wizardly to extend the power of a single voice? Nope. I mean, I can write this little essay, and maybe my fifty readers will like it, but those fifty people and everyone they’ll share it with already agree with me.

But about twenty miles from here, there are a couple of dozen people who literally hold the fate of our political conversation in their hands. In fact, it’s been in their hands for quite some time now, and they’ve made decisions which, only in retrospect, appear to have been disastrous for our nation’s politics.

At Facebook, the News Feed is the main stream of information that people see when they use the service. It has become the single most important source of news and conversation for many if not most Americans. It is designed to show people information that they want, which largely means showing people what they already agree with, from people who they already are inclined to sympathize with.

At Google, the search results page answers billions of queries each day, from billions of people. The results are carefully shaped not just with regard to each query, but as much as possible conformed for the particular user, so that the user sees results they are more likely to want to click, which in essence means showing them information they already agree with.

I’m not the first to note that the creation of these echo chambers only serves to reinforce existing biases, and isolate people from diverse opinions that could broaden their horizons and enrich our society. I might be among the first to charge that the product managers who now lead Facebook News Feed and Google Search are failing at their jobs.

On its face that’s a ridiculous statement, as we are talking about two of the most successful products in history, literal world changers. And who could argue against the general strategy of conforming experience to user tastes? But there comes a time in the life cycle of even massively successful products, when the product has attained a use and effect that were never anticipated through all of the prior success. Product managers who do not grapple with what their products have become, in all dimensions, are not doing their jobs well.

News Feed and Search are unique in the landscape of all products. These are no longer simply things that people use, and therefore need to be designed to be as pleasant and popular as possible. These products now form the infrastructure of political conversation, they have become the backbone of our polity, they are the means by which citizens of our nation engage with each other on the essential ideas of community. The success of these products must now be judged on how well they serve beneficial outcomes in our society, especially our politics.

There are plenty of people at Facebook and Google who are deeply invested in denying this responsibility, which is so self-evident to all of the rest of us mere users. They would like to say that their products are designed to be “neutral,” to simply follow algorithms that have no sense of society or humanity. They want to hide their power behind obfuscating explanations of math and probability.

Some of this may be a difference in perspective. Some of this may be benign short-sightedness. But some of it is moral cowardice. I hate to make such an inflammatory charge, but when you have the ability to shape a product in a way that would reduce the likelihood of a fascist from taking the reins of a country with the firepower to end life as we know it, and you deny that you have this power, I have a hard time calling this anything other than what it is.

Facebook and Google know that their products contribute to a stifled political conversation that only hardens lines of hate and allows well-meaning people to isolate themselves in their own safe spaces. Will they continue to build their products in a way that divides our society? Or will they take real moral responsibility for how their products shape our political conversation, and make their products a conduit for uncomfortable ideas that could improve our world? Will they break down the barriers between hardened positions, expose ignorance to truth, measure hatred and inject love? Or will they claim that these goals are too soft, and anyway achieving them is too hard?

When Philip Morris discovered that their product was killing their customers, they hid the evidence for as long as they could, and they denied the truth even after it was apparent to everyone else, all so they could squeeze out the last dollars from their death-dealing empire. When Coca-Cola realized that sugary drinks were contributing to unprecedented rates of obesity, they diversified their product lines to include healthy drinks as well as sugar bombs – not exactly admirable, but at least preparing for a shift where people who could watch out for themselves would continue to contribute to the company’s bottom line. At this point, I would be okay with lesser evils, but I would prefer to see moral courage. Product managers at Facebook News Feed and Google Search: Do Your Jobs.

black and blue

~ ginsu ~ 1 Comment

I began writing this in the morning of July 7, disheartened by the killing of Philando Castile, struggling to make sense of this ongoing slaughter of black lives. I have a lot of respect for law enforcement as well, and I wondered how we will rebalance the scales of justice to make a better world. I sat down to write that Black vs Blue is a literal tradeoff of lives, not just a one-sided injustice, and it will only get worse if both sides cannot explicitly acknowledge this. Then I paused my writing, checked the news, and saw that it’s already gotten worse. I won’t change anything I’ve written so far, and I’ll call out where I paused and restarted …

#BlackLivesMatter is the apotheosis of hashtags, arguably the mark of separation between the Internet as merely united information and the Internet as truly united culture. Uniting the world doesn’t happen with the ties that bind – those ties were always there, if we had the eyes and heart to see and feel them. Unity comes with a force implacable enough to sink us if we can’t acknowledge what’s been there all along.

In dusty eons past, the world’s knowledge was stored on fragile dried plant pulp, stacked up in schools, libraries, churches and palaces, separated by uncrossable oceans, vast and hostile distances. Fast-forward a few centuries and everything is instantly connected, world knowledge is united, human progress progresses in a coordinated manner that, if not moving in lockstep yet, fairly resembles loosely bound rafts strapped together on a common ocean. At the bleeding edge of the modern evolution of common knowledge, participants in the exchange of information must form a common culture, at least to the extent they are forced to deal with the vast gulf that yet remains between us all despite swimming in the same waters. #BlackLivesMatter is the iceberg that must be reckoned with, ignore it at the peril of having our boats wrecked by the rocky mass underwater. It is a call for the absorption of information at first, but more deeply it is a call to grapple with history, with the experiences of another, with the immediate reaction and grief, the defensiveness and guilt, the raw emotions of others all paddling in different directions to pull our own rafts out of the way of the iceberg, futilely ignoring the unbreakable ropes that bind us together.

… I am hoping that high-falutin’, fantastically florid introduction has gotten rid of the idiots, before I get to the point, if there is one. Because I’m not engaging with the #AllLivesMatter or #WhiteLivesMatter crowd. As entertaining as it can be to methodically unpack the ignorance, fear, guilt and willful blindness of those reactionary hashtags, they are nothing more than the impotent flags flown atop the largest rafts whose inhabitants see the iceberg and ignore it, not seeing that the ropes between the rafts are indeed unbreakable, and no hull or armor will protect their own raft from the sharp rocks beneath the water. #AllLivesMatter and #WhiteLivesMatter are false flags to be ignored. #BlackLivesMatter is the iceberg.

But #BlueLivesMatter is not so easily dismissed, although its standard bearers can seem similar to the false flag brigade, in being across such a great divide of cultural understanding that the good instinct to dismiss false flags casts its shadow over this hashtag as well. And a call to value police lives seems absurd as we watch, over and over again, the deaths of Eric Garner, Michael Brown, Laquan McDonald, Akai Gurley, Tamir Rice, Walter Scott, Freddie Gray, Alton Sterling, Philando Castile … at this point the despair of listing so many names is swamped by the certainty that we need wait only weeks to add another to the list. The volume of clear police crime and misbehavior may lead one to believe that #BlueLivesMatter only defends individual and institutional racism, corruption, appalling training and criminal collusion.

Hashtags, like people, aren’t monolithic. Different people mean different things when they post, tweet, instashare, snapblog, or whatever we do to cry our digital tears in the electronic ether. But in the best interpretation of #BlueLivesMatter, it really does spring from a concern for life, and there are the utmost consequences for the lives that will be affected when #Black meets #Blue. For #BlackLivesMatter to be given its full due, #BlueLivesMatter will have to make the ultimate sacrifice in some instances.

The question here is how to share the burden of mistakes. On my best days, I can imagine a world without racism (or at least, without the level of racism that directly or indirectly leads to death), as idealistic and distant as that world may be. But I can’t imagine a world without mistakes, which are so fundamental to the human condition, and necessary to human progress, that we ought to fear a world without mistakes more than one with racism.

So in a world with mistakes, where we also have deadly weapons and criminals, unless we want criminals to be able to kill with impunity, we must have police or accept vigilantes. And if you have police in this world, some of them must be armed, and some of them will make mistakes. (I suppose we can imagine a world without deadly weapons, or one without criminals, but if all we do is imagine a world without problems, there isn’t much to discuss.) So the police will make mistakes involving deadly weapons, and someone will die as a result. Who should it be, an innocent suspect, or an innocent police officer or bystander?

#BlueLivesMatter says that the innocent suspect must die. This sounds horrible, but it isn’t hard to envision math that makes this seem rational. Say there are 1000 life-or-death suspect vs police confrontations per year. Say that 10% of those confrontations will end in a mistake, so we count up the number of lives at stake as something like:

  • 100 suspects
  • 300 police officers
  • 500 bystanders

Again, we are assuming that all of these actions will be a police mistake, but we’re not yet deciding where the burden of mistakes lies. Let’s say that the mistake in every instance is that the police show up and start shooting – they shoot all of the suspects (100), 10% of the bystanders (50), and 5% of the cops (15). That’s 165 dead, all of them innocent.

Now let’s say that instead, the mistake is that the cops never shoot. They’re unarmed, or even if they are armed, they never unholster. But again, this is a fatal mistake. In this type of mistake, the suspects are not innocent, so all of the cops and bystanders die. That’s 800 innocent dead.

Obviously, this is not the actual data – the point here isn’t to examine the data, but to question the concept. If #BlueLivesMatter is saying that they’d rather see innocent suspects die than innocent police and bystanders, that is not an argument that can be dismissed conceptually. It is at least worthy of discussion. An informed discussion would require actually examining the data. It is wrong to say that #BlueLivesMatter is a conceptually nonsensical response to #BlackLivesMatter, as we can say about #AllLivesMatter and #WhiteLivesMatter. #BlueLivesMatter has conceptual integrity, but lacks data.

Coming out of the conceptual world and back to reality: much of the data that would be needed is never collected, and probably can’t be reliably collected. How exactly do you count lives saved? How exactly do you count mistakes made by police inaction? In addition, the sad reality is that many elements of police culture are devoted to hiding what little data we might have. And then the truly despairing reality is that in the vast majority of innocent suspects who are killed, the victim is black, and some of those killings are less properly characterized as mistakes than as manifestations of direct and institutional racism.

Nevertheless, conceptual integrity matters. #BlackLivesMatter demands that …

This is where I paused. Check news. Eleven officers shot, four confirmed dead. What the fuck was I doing in the conceptual world? What the fuck is happening in the real world? Where the fuck is all this heading?

#BlackLivesMatter demands that we reduce the slaughter of innocent black lives, without explicitly recognizing the treacherous mass of iceberg beneath the surface. In a somewhat more perfect world, one without racism but still with crime and guns and deadly mistakes, to reduce the killing of one set of innocents will increase the killing of another another set. We don’t know if the tradeoff is 1-for-1, 10-for-1, or 1-for-10; we’ve never seen the data and may not be able to get it. But if police officers are trained to be slower on the draw, to err on the side of caution, that means that they must be trained such that more police officers will die. Maybe it’s worth it. Maybe 1000 innocent black lives will be saved in return for 1 police officer’s death. But what do you say to that one police officer’s family?

It’s easy to object, “Well how about we just get rid of racism in the police force first, and then go from there? No one here is asking cops to die, dummy!”

Get rid of racism in the police force? Do you have a formula for that? How about we get rid of racism in the classroom, in the workplace, in the legislature, in the entire world while we’re at it? We don’t know how to do this within a group of kindergarten kids, how are we going to do this for the more than 1 million police officers in the United States? Why do police officers bear the responsibility for curing racism more than any other group, how can we ask them to sacrifice even one of their own?

This is not to say that we can do nothing. Of course we can. Of course there is better training, better hiring, better testing, better evaluation, better management. And when there is still racism, there can be better recognition, better correction, better enforcement and true and swift punishment. Nevertheless the package of all of those things is going to mean, in addition to less racism and sundry other benefits: more caution. Saving innocent black lives is about changing the balance of mistakes. It is about erring on the side of caution so that fewer innocent black lives are ended – and it therefore also means that some more innocent blue lives must end.

This is a mortal conflict. It is black lives against blue lives, there must be a tradeoff. I don’t know where the tradeoff is, and although we don’t have sufficient data, I feel we will need to act without all of the data in order to prevent greater carnage. Because if advocates, thought leaders, legislators, and well-meaning citizens on both sides don’t admit this is a mortal trade-off, and make the hard decisions to force those trade-offs, then the formula will continue to get worse. More lives will be lost on every side.

More innocent black men will die. In response, there will be more outrage. We no longer live in a world where this outrage can be suppressed by a combination of obfuscation, brutality and social pressure. It will only grow until there is outright war in the streets. And here is where we are now. Two innocent black men dead this week. Four innocent blue lives ended in Dallas today. This isn’t slowing down, it isn’t blowing over, it isn’t getting better.

That’s all I can manage right now, an uncharacteristically unedited dump of thoughts. #Black vs #Blue is an ever-growing river of red.