This Week’s News & Human Machine Update

Hello, again.

We’ve been hard at work creating an in-real-life Human Machine experience, and we’re super excited to share it with you in this email. We’ll get to that in a little bit, but first a quick catch-up of this week’s news in ethical technology.

It’s Time to Breakup Facebook by Chris Hughes in the New York Times

The piece that everyone has been talking about, and for good reason. We’ve heard calls from different corners of the political world to breakup large tech companies, the most prominent coming from Democratic Presidential candidate Senator Elizabeth Warren. This one though comes from inside of Facebook’s house. Chris Hughes, co-founder of Facebook, makes what is maybe the most articulate and detailed argument for why and how Facebook should be broken up. The whole piece if worth reading, but this quote stood out to me:

“I’m disappointed in myself and the early Facebook team for not thinking more about how the News Feed algorithm could change our culture, influence elections and empower nationalist leaders.”

You can read the entire piece in the New York Times here.

Google Enters the Privacy Ring

This week, Google hosted their annual developer conference and largest event of the year, Google I/O. During the event’s keynote, they announced new products and even some rebranding, but by far the biggest news was their heavy emphasis on privacy and user protection.

Google has been somewhat pushed into this position due to recent scandals and Apple’s public focus on privacy as a competitive advantage. Google’s claim is that they can provide superior machine learning and artificial intelligence while also protecting user privacy. This is a recent development, due to their newly gained ability to compress their machine learning models and store them locally on your device, bringing new levels of speed and security for users.

You can read more about Google’s announcements in this round-up by Ina Fried on Axios.

The Next Phase for Human Machine

This week, in collaboration with Creative Mornings and powered by Civic Hall, we hosted a “Creating Ethical Products” workshop. During the workshop, we covered the ethical areas most companies should focus on, and explored a few of those in detail using ethical games.

We’re excited to host more workshops, talks, and events in the future so stay tuned. In the meantime, you can get a sneak peak of the workshop by downloading our takeaway guide.

Who lives, who dies, who tells the self-driving car?

Who lives, who dies, who tells the self-driving car?

Researchers at MIT recently completed a massive survey to better understand how we would prefer for self-driving cars to decide who gets to live or die given a situation where collision is inevitable, bringing the “Trolley Problem” into the real world.

The Trolley Problem is a classic ethical dilemma that you've likely heard of. It goes like this: imagine that you are standing on a train platform, and an out of control train is speeding towards a group of 5 people who do not have the ability to escape. Next to you is a switch that would redirect the runaway train towards a different track which only has one person in harm's way. What do you do? Do you look away and take no action, causing the 5 people to die, or do you take action by flipping the switch and killing the single individual on the other track, therefore taking an active role in their death? Which of these is the more ethical option?

To make it more complicated, imagine if those 5 individuals are criminals guilty of murder, and the one individual is a child. Does that change your decision?

This is exactly what MIT researchers wanted to discover. Self-driving cars will have to make these decisions quite frequently, and they'll end up doing whatever we program into them. So would it be ethical for us to program a value system that ranks the importance of different human beings? Should we leave the decision engine neutral and risk killing a family in order to save a cat?

With over 40 million survey submissions, it's clear that we have strong preferences, some of which are fairly grim. Starting from the bottom, research participants ranked cats and criminals as the least important obstacles to save. Right above that, dogs (take that, cat people.) Older and even larger adults also ranked near the bottom. Near the top of the prioritized list were children and pregnant women.

Whether any sort of ranking system like this exists today is unclear. Tesla, Alphabet, and Apple continue to develop their self-driving car technology, and one can only presume that these kinds of discussions are being had.

If a human driver faced the same situation, it's clear that the decision would be made in a split second, without any time to really have an internal debate over which life matters more. None of the choices would be ideal. But in creating autonomous vehicles, we do have the time to consider these problems, and whatever we decide will be programmed into millions of cars. Much like the human driver, there is no clear right answer.

Teams should take a hard look at the technical limitations that could be used in order to inform that decision. For example, a self-driving car does not have the technical ability to distinguish between an adult with a criminal record and one without. So what can these cars actually distinguish between, and should we use that data to inform whether a car decides to protect the object in front of the vehicle or behind the wheel? To use that data or not use that data, that is the question.

Apple Goes All In

Apple Goes All In

For the last 5 years, as companies like Google and Samsung have caught up on the design front, Apple has been staking out new territory to differentiate itself. Its walled garden approach and “it just works” ecosystem has long been a selling point for Apple, and a point of criticism from its competitors. But in this new era of machine learning and data-as-a-business-model, Apple’s philosophy may start paying off for a whole new reason besides efficiency, and it’s never been as clear as it was this week. 

Privacy snuck into Apple’s value proposition as recently as 4 or 5 years ago, with the introduction of voice assistants on mobile phones. As more companies like Google and Facebook began to actively collect user data and build products that would monetize that data, Apple decided to run in the opposite direction. Personal data would be processed on device, Apple said, not at their data centers. Your data would remain your data. Apple even installed a secure chip inside iPhones and Macs in order to store highly sensitive information like your fingerprint, facial map, and credit cards.  

The highest profile case to put Apple’s dedication to privacy to the test came in the aftermath of the San Bernardino shooting. The FBI demanded that Apple unlock an iPhone belonging to one of the attackers, and Apple refused. This lead the federal government to spend close to $1.3 million dollars in order to hack their own way into the device. 

Ever since then, Apple has repeatedly rejected the call to build “back doors” into their devices to make it easier for law enforcement to access them. But they didn’t stop there. Apple pushed to encrypt more data on your device, and even introduced Face ID in order to increase the difficulty of unlocking a device. 

During that time, the concept of privacy began showing up on Apple’s product pages and keynote announcements. Which leads us to the past week. Two of Apple’s most prominent executives, CEO Tim Cook and CDO Jony Ive both spoke up publicly about the importance of user privacy and the consideration of product ethics during the design and development process. 

“If you’re creating something new, it is inevitable there will be consequences that were not foreseen — some that will be great, and then there are those that aren’t as positive,” Ive told the Financial Times. 

“There is a responsibility to try and predict as many of the consequences as possible and I think you have a moral responsibility to try to understand, try to mitigate those that you didn’t predict.

If you genuinely have a concern for humanity, you will be preoccupied with trying to understand the implications, the consequences of creating something that hasn’t existed before.”

This is easily the most public expression of support for the consideration of product ethics from one of tech’s most important leaders. On the other side of the globe, Tim Cook was making a similar point. 

“At every stage of the creative process, then and now, we engage in an open, honest, and robust ethical debate about the products we make and the impact they will have. That’s just a part of our culture,” Cook told the Debating Ethics: Dignity and Respect in Data Driven Life conference in Brussels. 

"We don’t do it because we have to, we do it because we ought to. The values behind our products are as important to us as any feature.”

Apple’s proactive approach is in stark contrast to the reactionary game of whack-a-mole that Facebook, Google, and Twitter are engaged in. And if the world’s most valuable company is taking ethical considerations into account during their product development process, why shouldn’t every company be doing the same?

Cook offered a pretty simplistic framework for their ethical discussions. "We aren’t absolutists, and we don’t claim to have all the answers. Instead, we always try to return to that simple question:  What kind of world do we want to live in.”

In his speech, Cook declared privacy to be a human right, and called on more governments to implement laws and regulations to protect user data. 

"This crisis is real. It is not imagined, or exaggerated, or “crazy.” And those of us who believe in technology’s potential for good must not shrink from this moment.”

As more scandals continue to embroil tech companies, and Apple continues to pull away as a leader in privacy and product ethics, the question becomes, will the rest of the industry catch up?

Do codes of ethics work?

Do codes of ethics work?

​Hola from the Cartagena International Airport! 

​​A recent study done by North Carolina State University set out to uncover whether codes of ethics were an effective method of helping software engineers make better, more ethical decisions in the product development process. 

​​The results showed that being exposed to a code of ethics did little to alter an engineer’s decision. 

​​​​“Despite stated goals, we found no evidence that the ACM (Association of Computing Mechanics) code of ethics influences ethical decision making.”

​​This outcome makes a lot of sense. Codes of ethics are like a compass, they won’t necessarily tell you exactly where to go and how to get there.  They can still be useful tools, but only if they’re widely accepted by the organization and regularly enforced when major decisions are being made. 

​​The study also found, but did not further explore, the fact that case studies and recent news of tech controversies had a bigger impact on developer behavior. Case studies are helpful because they show the impact of unethical decisions. Where a code may feel academic, case studies, workshops, and ethical sprints bring these kinds of decisions to the real world, and highlight the actual impact they have on users and society at large. 

​​This study validates a lot of what we believe at human machine. Technology ethics are not theories, but real practical considerations that we should make during every point of product development. 

​​​​You can can read the full study here: https://people.engr.ncsu.edu/ermurph3/papers/fse18nier.pdf

---

​​Here’s what else you should know…

“Instagram has a massive harassment problem.” in The Atlantic https://www.theatlantic.com/technology/archive/2018/10/instagram-has-massive-harassment-problem/572890/

​​

What we're working on.

What we’re working on.

So far, the human machine newsletter has been dedicated to covering recent tech news within the context of the ethical implications of that tech. We believe that it’s important for everyone to understand the rapid developments taking place around them through the lens of their ethical implications. But our goal at human machine goes beyond that than that. We want to help designers, developers, entrepreneurs, investors, and other tech leaders to make better ethical decisions before their creations have unintended societal consequences(oneswe are now all too familiar with). 

Within the boundaries of technology ethics exists a vast landscape of decisions, all of them shades of gray. We’re not here to give you the answer, and we’re not here to define what is right or wrong. We’re simply here to help you ask better questions, and make better decisions. Outside of those boundaries is a dark world of unethical behavior from exploitation, unequal treatment based on gender, sexuality, race, religion, ethnicity, or ability, harassment and violence, and strategic & purposeful deception. We’re here to call those things out and help you protect your users, employees, and products from those forces. 

Ask better questions. Make better decisions. 

To do that, we’re developing a framework to help you gain visibility into all of the different ways a product or service could have unintended negative consequences on individuals and society as a whole, and whether that product or service could be easily co-oped by those looking to do harm. The framework will take many different forms: a canvas, a white paper, ethical sprint workshops, individual exercises, and so on. The goal is to normalize this conversation to the point where ethical considerations during the development of a product feel simple and necessary. 

The framework is still under development, but here’s a sneak peek into two key principles we’ve identified and sub-categories within them. 

USER & EMPLOYEE SAFETY

Ethical products are designed with user safety in mind by employees who themselves work in a safe environment. User safety can mean many things, but mainly, our framework will address:

• Data security

• Protection from harassment and hate speech

• Content moderation policies

• Accessibility

As for employee safety, our framework includes the need for: 

• Codes of conduct

• Accessibility, diversity, and inclusion

• Equity

NET POSITIVE SOCIETAL IMPACT

A decision is made every single day whether to built or fund a new feature or product. Those decisions are usually made based on whether that product or feature will have a positive impact on the bottom line. Our framework encourages organizations to also consider the societal impact it would have, including: 

• Impact on the environment

• Long-term user behavior impact

• Large scale societal impact

• Protection from bad actors 

The framework is a work in progress, and we’re excited to share more with you as we develop it.