The Merriam-Webster online dictionary among several options gives this definition for human: “[…]representative of or susceptible to the sympathies and frailties of human nature human kindness a human weakness.”
Then there is humanity which can either confer either the collective of humans, or “[…]the fact or condition of being human; human nature,” or benevolence as in compassion and understanding. For the latter, it seems that we are the eternal optimists when it comes to describing ourselves. Hence, we often refer to the humanity of man as one of our most redeeming traits. At the same time, if we query human nature we can get, “[…]ordinary human behavior, esp considered as less than perfect.” This is a diplomatic way of acknowledging that flaws are a characteristic of our nature. When we talk about our humanity, we presumptively leave out our propensity for greed, pride, and the other deadly sins. We like to think of ourselves as basically good.
If we are honest with ourselves, however, we know this is not always the case and if we push the issue we would have to acknowledge that this not even the case most of the time. Humanity is primarily driven by the kinds of things we don’t like to see in others but rarely see in ourselves. But this is supposed to be a blog about design and tech, isn’t it? So I should get to the point.
A recent article on the blog site QUARTZ, Sarah Kessler’s article, “Algorithms are failing Facebook. Can humanity save it?” poses an interesting question and one that I’ve raised in the past. We like to think that technology will resolve all of our basic human failings—somehow. Recognizing this, back in 1968 Stewart Brand introduced the first Whole Earth Catalog with,
“We are as gods and might as well get good at it.”
After almost 50 years it seems justified to ask whether we’ve made any improvements whatsoever. The question is pertinent in light of Kessler’s article on the advent of Facebook Live. In this particular FB feature, you stream whatever video you want, and it goes out to the whole world instantly. Of course, we need this, right? And we need this now, right? Of course we do.
Like most of these wiz bang technologies they are designed to attract millennials with, “Wow! Cool.” But it is not a simple task. How would a company like Facebook police the potentially billions of feeds coming into the system? The answer is (as is becoming more the case) AI. Artificial Intelligence. Algorithms will recognize and determine what is and is not acceptable to go streaming out to the world. And apparently, Zuck and company were pretty confident that they could pull this off.
Maybe not. Kessler notes that,
“According to a Wall Street Journal tally, more than 50 acts of violence, including murders, suicides, and sexual assault, have been broadcast over Facebook Live since the feature launched 13 months ago.”
Both articles tell how Facebook’s Mark Zuckerberg put a team on “lockdown” to rush the feature to market. What was the hurry, one might ask? And Kessler does ask.
After these 13 months of spurious events, the tipping point came with a particularly heinous act that ended up circulating on FB Live for nearly 24 hours. It involved a 20-year-old Thai man named Wuttisan Wongtalay, who filmed himself flinging his 11-month-old daughter off the side of a building with a noose around her neck. Then, off-camera, he killed himself.
“In a status update on his personal Facebook profile, CEO Mark Zuckerberg, himself the father of a young girl, pledged that the company would, among other things, add 3,000 people to the team that reviews Facebook content for violations of the company’s policies.”
Note that the answer is not to remove the feature until things could be sorted out or to admit that the algorithms are not ready for prime time. The somewhat surprising answer is more humans.
Kessler, quoting the Wall Street Journal article states,
“Facebook, in a civic mindset, could have put a plan in place for monitoring Facebook Live for violence, or waited to launch Facebook Live until the company was confident it could quickly respond to abuse. It could have hired the additional 3,000 human content reviewers in advance.
But Facebook ‘didn’t grasp the gravity of the medium,’ an anonymous source familiar with Facebook’s Live’s development told the Wall Street Journal.”
Algorithms are code that helps machines learn. They look at a lot of data, say pictures of guns, and then they learn to identify what a gun is. They are not particularly good at context. They don’t know, for example, whether your video is, “Hey, look at my new gun?” or “Die, scumbag.”
So in addition to algorithms, Zuck has decided that he will put 3,000 humans on the case. Nevertheless, writes Kessler,
“[…]they can’t solve Facebook’s problems on their own. Facebook’s active users comprise about a quarter of the world’s population and outnumber the combined populations of the US and China. Adding another 3,000 workers to the mix to monitor content simply isn’t going to make a meaningful difference. As Zuckerberg put it during a phone call with investors, “No matter how many people we have on the team, we’ll never be able to look at everything.”[Emphasis mine.]
So, I go back to my original question: We need this, right?
There are two things going on here. First is the matter of Facebook not grasping the gravity of the medium (which I see as inexcusable), and the second is how the whole thing came around full circle. Algorithms are supposed to replace humans. Instead we added 3,000 more jobs. Unfortunately, that wasn’t the plan. But it could have been.
Algorithms are undoubtedly here to stay, but not necessarily for every application and humans are still better at interpreting human intent than machines are. All of this underscores my position from previous blogs, that most companies when the issue is whether they get to or stay on top, will not police themselves. They’ll wait until it breaks and then fix it, or try to. The problem is that as algorithms get increasingly more complicated fixing them gets just as tricky.
People are working on this so that designers can see what went wrong, but the technology is not there yet.
And it is not just so that we can determine the difference between porn and breastfeeding. Algorithms are starting to make a lot of high stakes decisions, like autonomous vehicles, autonomous drones, or autonomous (fill in the blank). Until the people who are literally racing each other to be the first step back and ask the tougher questions, these types of unanticipated consequences will be commonplace, especially when the prudent actions like stop and assess are rarely considered. No one wants to stop and assess.
Kessler says it well,
“The combination may be fleeting—the technology will catch up eventually—but it’s also quite fitting given that so many of the problems Facebook is now confronting, from revenge porn to fake news to Facebook Live murders, are themselves the result of humanity mixing with algorithms.” [Emphasis mine.]
We can’t get past that humanity thing.