Is AOC right about AI?

Conservative Twitter is up in arms today over Rep. Alexandria Ocasio-Cortez saying at an MLK Day event that algorithms are biased. (Of course “bias” has been translated into “racism.”) The general response from the right has been, “What a dumb socialist! Algorithms are run by math. Math can’t be racist!” And from the tech experts on Twitter: “Well, actually….”

I have to put myself in the latter camp. Though I’m not exactly a tech expert, I’ve been researching the impact of technology like AI and algorithms on human well-being for a couple of years now, and the evidence is pretty clear: people have bias, people make algorithms, so algorithms have bias.

When I was a kid, my dad had this new-fangled job as a “computer programmer”. The most vivid and lasting evidence of this vocation was huge stacks of perforated printer paper and dozens upon dozens of floppy disks. But I also remember him saying this phrase enough times to get it stuck in my head: “garbage in, garbage out.” This phrase became popular in the early computer days because it was an easy way to explain what happened when flawed data was put into a machine – the machine spit flawed data out. This was true when my dad was doing…whatever he was doing… and when I was trying to change the look of my MySpace page with rudimentary HTML code. And it’s true with AI, too. (Which is a big reason we need the tech world to focus more on empathy. But I won’t go on that tangent today.)

When I was just starting work on my book, I read Cathy O’Neil’s Weapons of Math Destruction (read it.), which convinced me beyond any remaining doubt that we had a problem. Relying on algorithms to make decisions for us that have little to no oversight and are entirely susceptible to contamination by human bias – conscious or not – is not a liberal anxiety dream. It’s our current reality. It’s just that a lot of us – and I’ll be clear that here I mean a lot of us white and otherwise nonmarginalized people – don’t really notice.

Maybe you still think this is BS. Numbers are numbers, regardless of the intent/mistake/feeling/belief of the person entering them into a computer, you say. This is often hard to get your head around when you see all bias as intentional, I get that, I’ve been there. So let me give you some examples:

There are several studies showing that people with names that don’t “sound white” are often passed up for jobs in favor of more “white-sounding” names. It reportedly happens to women, too. A couple of years ago, Amazon noticed that the algorithm it had created to sift through resumes was biased against women. It had somehow “taught itself that male candidates were preferable.” Amazon tweaked the algorithm, but eventually gave up on it, claiming it might find other ways to skirt neutrality. The algorithm wasn’t doing that with a mind of its own, of course. Machine-learning algorithms, well, learn, but they have to have teachers, whether those teachers are people or gobs of data arranged by people (or by other bots that were programmed by people…). There’s always a person involved, is my point, and people are fallible. And biased. Even unconsciouslyEven IBM admits it. This is a really difficult problem that even the biggest tech companies haven’t yet figured out how to fix. This isn’t about saying “developers are racist/sexist/evil,” it’s about accounting for the fact that all people have biases, and even if we try to set them aside, they can show up in our work. Especially when those of us doing that work happen to be a pretty homogeneous group. One argument for more diversity in tech is that if the humans making the bots are more diverse, the bots will know how to recognize and value more than one kind of person. (Hey, maybe instead of trying to kill us the bots that take over the world will be super woke!)

Another example: In 2015, Google came under fire after a facial recognition app identified several black people as gorillas. There’s no nice way to say that. That’s what happened. The company apologized and tried to fix it, but the best it could do at the time was to remove “gorilla” as an option for the AI. So what happened? Google hasn’t been totally clear on the answer to this, but facial recognition AI works by learning to categorize lots and lots of photos. Technically someone could have trained it to label black people as gorillas, but perhaps more likely is that the folks training the AI in this case simply didn’t consider this potential unintended consequence of letting an imperfect facial recognition bot out into the world. (And, advocates argue, maybe more black folks on the developer team could have prevented this. Maybe.) Last year a spokesperson told Wired: “Image labeling technology is still early and unfortunately it’s nowhere near perfect.” At least Google Photos lets users to report mistakes, but for those who are still skeptical, note: that means even Google acknowledges mistakes are being – and will continue to be – made in this arena.

One last example, because it’s perhaps the most obvious and also maybe the most ridiculous: Microsoft’s Twitter bot, Tay. In 2016, this AI chatbot was unleashed on Twitter, ready to learn how to talk like a millennial and show off Microsoft’s algorithmic skills. But almost as soon as Tay encountered the actual people of Twitter – all of them, not just cutesy millennials speaking in Internet code but also unrepentant trolls and malignant racists – her limitations were put into stark relief. In less than a day, she became a caricature of violent, anti-semitic racist. Some of the tweets seemed to come out of nowhere, but some were thanks to a nifty feature in which people could say “repeat after me” to Tay and she would do just that. (Who ever would have thought that could backfire on Twitter?) Microsoft deleted Tay’s most offensive tweets and eventually made her account private. It was a wild day on the Internet, even for 2016, but it was quickly forgotten. The story bears repeating today, though, because clearly we are still working out the whole bot-human interaction thing.

To close, I’ll just leave you with AOC’s words at the MLK event. See if they still seem dramatic to you.

“Look at – IBM was creating facial recognition technology to target, to do crime profiling. We see over and over again, whether it’s FaceTime, they always have these racial inequities that get translated because algorithms are still made by human beings, and those algorithms are still pegged to those, to basic human assumptions. They’re just automated, and automated assumptions, it’s like if you don’t fix the bias then you’re automating the bias. And that gets even more dangerous.”

(This is the “crime profiling” thing she references, by the way. I’m not sure where the FaceTime thing comes from but I will update this post if/when I get some context on that.)

Update: Thanks to the PLUG newsletter (which I highly recommend) I just came across this fantastic video that does a wonderful job of explaining the issue of AI bias and diversity. It includes a pretty wild example, too. Check it out.

Advertisements

On empathy, endorsement, and what happens next

I walk a little over a mile to and from work each day, and I usually spend it listening to podcasts, or to books on Audible. After more than a year of this, I really look forward to certain days when I know certain podcasts will have a new episode out. Note to Self is one of them. I love Manoush Zomorodi’s style of reporting on technology and how it affects our lives, and I love how she’s styled herself as a guide to “our accelerating world.” Because wow, yes, is it ever accelerating.

Note to Self is often about technology in a technical sense, but the show also takes occasional detours into the psychology of how we interact with tech. This, of course, is my favorite thing to write and read about. So I was really excited when Zomorodi recently interviewed Dylan Marron. He’s a progressive YouTuber and writer who also has a new podcast, called Conversations With People Who Hate Me. It’s pretty much exactly what it sounds like. Remember when Lindy West called her troll and it became a viral This American Life episode? Marron does a similar thing on each episode of Conversations. He talks to the people who profess to think he’s the scum of the earth, and tries to find out why.

This is something I’m going to write more about soon – podcasts and radical empathy – but for the purposes of this blog post I really want to focus on one thing Marron said during this Note to Self episode: empathy doesn’t mean endorsement. This is a fact that’s become so obvious to me, I think I forget to enunciate it to others when I talk about empathy. I’ve never found such a succinct way of saying it, either. But it’s absolutely the truth: sitting down and listening to someone does not necessarily mean validating them, and it definitely doesn’t mean agreeing with them. It’s just…acknowledging them. Taking their perspective.

That can feel a little scary. I know that I have had experiences in which I read something written by someone with vastly different views from my own and as I prepare to put myself in their shoes I think, what if I can’t get back out? What if they convince me? But things don’t really happen that abruptly, most of the time. We make our decisions and create our ideologies based on a mix of experiences and information, and it all sort of flows together and tries to balance itself out, rarely truly solidifying into one thing. What I mean is, we’re always learning, always changing our minds a little bit, even if we don’t always notice it, or want to.

I thought about this concept a lot as I watched the recent Alabama election unfold. Everyone around me kept asking, “How could these people vote for a pedophile?” I can’t say the answer is the same for everyone who voted for Roy Moore, but I can say pretty confidently that many of them did so because they didn’t believe what they heard about him. Or, they only believed parts of what they heard about him.

Brian Resnick has a great piece about this up at Vox. He interviewed a lot of Moore supporters in Alabama before the election, and reading this piece, I feel like I can really empathize with these people. Trying to put myself briefly in their shoes, I feel afraid, I feel disappointed, I feel betrayed. This is something I tried to do when reading story after story about Trump supporters last year as well. And I think it’s a worthwhile practice. But the part that nobody really seems to talk about is… then what?1

What do I do with this information? What do I do with the fact that people of all parties and ideologies cling to confirmation bias and “motivated reasoning?” Well, it’s made me feel a little bit less hopeless about change, for one thing. That might seem counter-intuitive, but knowing that we’re all susceptible to this, and witnessing people have conversations about it that don’t end in name-calling or fist fights, is encouraging. It also helps me feel less angry, which is no small thing. Over the past couple of years I’ve found anger to be less and less useful for me, at least on a personal level. Being mad at friends or family or strangers who did something I see as wrong doesn’t actually accomplish anything for me, except raising my blood pressure. When I understand their points of view a bit better, I can take some of the emotion out of my reaction to them. And if we’re both on the same page about that, we can have a conversation, and figure out where we agree. And sometimes… sometimes, one or both of us can bend a little. Without the pressure to immediately admit or agree to anything, this can feel a lot easier.

There’s one major caveat to all of this. And it’s never far from my mind when reading and listening to these conversations. This isn’t just about liberals learning to empathize with conservatives. There’s a lot going on in the other direction as well. And, especially after the election of Doug Jones over Roy Moore in Alabama, it’s way past time to start asking people to empathize with another group who doesn’t get nearly enough attention despite their huge impact and disproportionate burden: black voters. Especially black women. It’s good that we’re talking about empathy so much, but we also need to be real with ourselves about who we reserve it for.

Empathy for the holidays

This post is probably coming too late for many of you who just finished celebrating American Thanksgiving, but luckily there are several more holidays to come this year. Depending on how you celebrate, that may also mean several more opportunities to exercise empathy with family, friends, and even coworkers!

A tangled ball of Christmas lights

In the lead-up to Thanksgiving this year I came across several pieces of advice that really helped me put my own anxieties about the holiday into perspective. In general I feel like I’ve learned a lot over the past year about how to truly empathize with others, both as a result of the research I’ve been doing for my book, and because of the particularly heated moment we’re currently living through in American culture. But I still learn things that surprise me.

While watching Instagram Stories the day before Thanksgiving, for example, I came across a list of tips for talking with families, from the organization Showing Up For Racial Justice. I thought I knew what to expect, but I was actually a little surprised by the first item on the list:

Listen mindfully before formulating a thoughtful response.

We talk a lot about listening to others’ opinions, especially when they differ widely from our own, but what does it mean to listen mindfully? I actually came to this post because it was shared by a celebrity, who added this question: Are you listening to answer, or to understand? Are you taking in everything the other person says just to help you formulate a response, or are you actually considering this person’s thoughts for their own sake?

It sounds so simple, but it’s actually not easy to do in the moment. I had to sit with myself and think about what I’m really doing when I listen to people with whom I disagree. I thought about one family member who likes to debate politics on Facebook and via email. I realized that when I’m reading her messages, I am often just sitting there planning how I’m going to respond. I wondered, if I read them more mindfully, could our conversations go more smoothly? Maybe, maybe not. But it seemed worth a try.

The rest of this list included things you might find obvious – asking questions, respectfully affirming differences, breathing. But there was another one that struck me:

Notice what is possible for you at this time – stretching into discomfort while also caring for yourself.

This notion that sometimes, if you don’t feel up to it, you can opt out of hard conversations, has been a tough lesson for me to learn. But ultimately it’s better for everyone. It’s hard to really show empathy to others in conversation, to listen mindfully, when you feel like you don’t have it all together yourself. Yes, it is important to “stretch into discomfort” in order to learn new things and understand others better, but there’s no rule that you have to do it at the expense of everything else.

Hopefully you find some benefit in these tips. I did, even though I didn’t end up needing to use them at Thanksgiving. But they feel like good pieces of advice even for everyday conversations, whether they are about race, politics, work, relationships, or even sports. It can sometimes be hard to understand how to work empathy into our everyday communications, but thinking about it as “mindful listening” might help!

A conversation with Dr. Laura Roselle

“If fear is so important as an emotion, then empathy must be important as well.”

We talk a lot about empathy in relationships, in classrooms, and in the media. But this so called soft skill is also finding more credence among scholars working on some of the biggest geopolitical questions facing the world.

Last fall, I interviewed Dr. Laura Roselle, a professor of political science and policy studies at Elon University (disclosure: she was my International Relations professor there) about how the field is increasingly embracing empathy. The traditional ways of looking at geopolitical communication, which tend to focus on the “rational actor” theory that individuals (and states) always make logical decisions, can’t necessarily explain some of the biggest events in recent history, from the Cold War to the 2016 election. So some scholars are pushing for a different approach.

I spoke with Dr. Roselle for this story for Quartz, which ran just before the election. I didn’t have a chance at the time to really dig into some of the changes she is seeing in the field. Below is the rest of our conversation, in Q&A form.

On Empathy [Kaitlin Ugolik]: It can seem a bit counterintuitive to talk about things like emotions and empathy in the International Relations and political science context, but is that beginning to change?

Laura Roselle: Yes. Some feminist scholars are saying there’s more than just power politics, more than just a masculine narrative, to any interaction. Power has traditionally been defined in the sense of getting somebody to do something they wouldn’t ordinarily do. But there’s a totally different definition that you could use, which is the ability to create consensus, which absolutely requires empathy.

OE: Can you talk a bit more specifically about these new strains of scholarship and how they incorporate empathy?

LR: There are at least two strains of writing within political science that are very interested in this notion of empathy. One is among feminist scholars, who are writing about empathetic cooperation in international relations. It’s this notion that the whole narrative of what IR is, is highly masculinized. Christine Sylvester championed this notion of focusing on empathetic cooperation. The idea is that if we’re going to be talking about power, how we define power is really important.

The other is that in the last decade, there’s increasingly been a focus on emotions in politics and the role of moving beyond this notion of a rational actor—talking about both affect and cognition, both emotions and logic. Some of that’s come from the inability of those who rely on this rational actor model to explain surprising things that maybe emotion would be involved in. There’s a new book our—Emotions in International Politics, edited by Yohan Ariffin—and there’s a section about empathy.

OE: What exactly is the role that empathy might play?

LR: It’s this notion of understanding empathy as a process and not strictly an emotion. If we understand empathy as a process that involves both cognition and affect, involves listening, then you can tie it to trust and to dialogue, which can then be tied into this notion of international politics. In order to understand any kind of cooperation, you need trust and dialogue, but essential to that is empathy.

If fear is so important—and we talked about fear this whole last election cycle—then empathy must be important as well.

OE: Where do you think we are in this evolution of thinking? What might be next?

LR: The field has been moving to encompass more than power politics and rational actor analysis for a while. Probably since the end of the Cold War, which was a pretty big thing they couldn’t explain. Since then, there’s been a move to get at more of that complexity, but the empathy part hasn’t been there until recently. This notion of emotions and empathy coming in is new, and seems like a really good strain to be incorporating in.

OE: Has this evolving way of thinking about IR changed how you teach?

LR: The way I put it to students now is: Which definition of power are you more drawn to? Making someone do something they don’t want to do, or creating consensus? Most people want consensus. Which is harder? Consensus. It’s harder and it requires things like empathy. Especially in a world that is becoming more horizontal, more flat, where you have people able to communicate across state boundaries in ways that are really interesting, this notion of cooperation and building consensus and being empathetic is probably our hope for the future.

Empathy for the hysterics

As I mentioned in a previous post, though we may want it to be – I believe empathy is not a magic bullet. It doesn’t miraculously make us stop fearing or hating or resenting each other. It’s a process, and in order for it to look like much more than listening, it has to be paired with things like love and compassion and action.

This morning, I read yet another call to empathy from a political expert. Attorney and Harvard Law professor Lawrence Lessig, who ran for president last year, published “Rules for a constitutional crisis” on Medium early this morning. He starts with personal history, explaining some of the impetus for his decision to become a lawyer, and goes on to argue for Congress’s vital role in addressing the “constitutional crisis” the country now faces under Trump. It’s really worth a read. But near the end, something confused me:

Because if America is to avoid slipping into civil war, the people we need to keep in focus are the people who elected Donald Trump. I get that the easy way to think and talk about those Americans is to call them racists, or sexists or idiots. No doubt there are some who are those (as there are some on the other side who are each of those things too). But it is neither true nor helpful to simplify this story into good versus evil. The citizens who elected Trump are not evil. And if America is going to survive this crisis, we need to convince them first that their President should not be President. We need to show them that their own values are consistent with ours, in this respect at least.

That won’t happen with hysterics. It won’t happen with violence. It won’t happen by behaving just as badly as Donald Trump is behaving. It will only happen if the opposition is, and seems, better than Trump. That is, if it inspires in all Americans—and especially a large swath of the supporters of Trump—a recognition of the ideals that we all know we are to embrace: the Constitution, the rule of law, and government officials who know their place within that system.

I’m not the only commenter who has asked, “What hysterics? What violence? Someone is ‘behaving just as badly as Donald Trump?’ Where!?” Though I actually agree with most of what Lessig says here, this kind of call for empathy reads to me more like a rebuke. It reads like it’s asking “the opposition” to practice empathy instead of calling for resistance, because the latter may be seen as “hysteria.” In my opinion, it creates a false equivalence that doesn’t seem helpful to anyone on either side (especially since there’s been little to no violence, it’s unclear who other than Trump – except maybe his core team – can be described to be behaving like Trump, and “hysterics” is an extremely loaded – and therefore not super useful – word).

I also think there’s an important difference between empathizing with someone and pandering to them. Am I saying that I believe the citizens who elected Trump are indeed evil? No! I’m simply saying that while empathy for them is important in understanding how we got here, I don’t believe that should trump (….) being honest about what’s happening, and how dire it might be. I also don’t believe, after many conversations with Trump voters, that liberals and/or Democratic congresspeople redirecting their energy from “hysterically” sounding the alarm to being more universally “inspiring” is going to change many minds. At least not right now. When people have been conditioned to see those who disagree as an enemy – and when this is the kind of spin those trying to win them over have to contend with – I’m not sure any change in tone is going to make a big difference. We know people tend to just get more entrenched in their own beliefs the more we try to convince them they’re wrong, anyway (though presenting good reasoning and facts can sometimes be persuasive).

What’s that saying about “drastic times and drastic measures?” Is it ever legitimate to get a little “hysterical?”

Interested in empathy? Check this out.

Today, the Huffington Post has a big list of ways to incorporate more empathy into your life this year . I’ve been ramping up my writing about empathy, but not here. I won’t be using WordPress anymore after next week. So if you want to read more of my writing about empathy, please subscribe to my newsletter here. You will get an intro email, and then just two emails each month. I would really hate to lose the conversations with all of you when I make the move, so I hope you’ll join me!

 

On empathy as a magic bullet

There’s been a lot of talk about empathy since the election. I have been thinking kind of obsessively about empathy for years, so it was natural for me to go there, but I was a little surprised to see some of the people and organizations that touted it as a way to cope with the election of Donald Trump and connect with family members and friends who voted for him.

I struggled with all of the competing messages. Who exactly deserves whose empathy? Are we all supposed to have it for one another? Or only some of “us” for some of “them?” Is everyone even working with the same definition of empathy?

Over the weekend, several people sent me this smart piece by Amanda Hess in the New York Times: Is ‘Empathy’ Really What the Nation Needs? She makes some really great points about who these directions for empathizing are coming from, and what interest they might have in people following their advice. Mark Zuckerberg, for example, has been a big proponent of empathy in the wake of the presidential election, and he wants us to practice it on Facebook, naturally. He also made these comments after many people questioned the role Facebook may have played in spreading conspiracy theories and fake news during the campaign. So yes, maybe we should be skeptical about calls for empathy.

Then there is Paul Bloom, psychology professor and author of the forthcoming book Against Empathy, who warns that trying to feel others’ pain in political and policy contexts actually distorts our reasoning and often ends up doing more harm than good. Yikes.

But what about at the personal level? In addition to the clarion call for empathy from various writers and technology executives, I’ve seen a lot of concern among friends, particularly on social media, that we have all failed at the important task of empathizing with other people as individuals. After  the election, many were shocked to find out that their grandmother or cousin voted for Trump out of a sense of frustration, despair, or in some cases racism, that hadn’t been apparent before. I know I personally looked at a couple of family members and friends and thought, Really? You? So in an effort to understand, I too looked to empathy.

But I have my own qualms with this approach, different from the concerns of writers like Hess and Bloom. My frustration comes from the fact that empathy is being discussed like a simple solution, a cure-all for bringing people together, a switch that can be flipped, when in fact, in practice, it seems to require a great deal of patience and trust-building. Anyone who’s ever had an argument with a stranger on social media has probably experienced this. When you don’t understand a person’s intentions (trolling? lacking context? just a mean person?) or anything at all about their life or history, it’s hard to connect on anything more than a superficial level. And at that level, do we really think we can put ourselves in the other person’s shoes? And even if we can, how likely is it that they will trust that we can, or that our intentions are good?

I don’t have the answers! I’m continually researching this. And I’m going through trial and error in my own life. I still don’t think it can hurt to empathize with others, even if it’s ultimately just an exercise for yourself. (Not that you should use other people to learn about yourself, but that if your attempt at empathy doesn’t seem well-received, you might learn something about the way you empathize, and your own intentions.) I also lean toward the belief that empathy is a vital part of long-term relationships with partners, family members and friends. But I do think it’s important that we ask these questions when something as complex and nebulous as empathy starts to seem like a buzzword, or a magical solution to political and policy issues.