All Tech Is Human’s 2021 Responsible Tech Year-in-Review
It’s been a big year for All Tech Is Human!
In 2021, we released three major reports (Improving Social Media, The Business Case for AI Ethics, and the updated Responsible Tech Guide), hosted two summits and multiple livestreams and mixers. We expanded and improved our Responsible Tech Job Board and built a new website, courtesy of Seattle design firm Artefact. Finally, we secured $650,000 in funding, giving us the resources that will enable us to execute our mission and further our goal of building the responsible technology pipeline. We also started a Slack group for the global Responsible Tech community that grew to over 2.2k members.
Our 2022 plans include a report from our Human Experience (HX) working group set for release in January, a Mentorship Program, and a major university ecosystem mapping initiative from our University Ambassadors, among other projects and events.
For All Tech Is Human's Responsible Tech Year-in-Review, here are some of the people, organizations, and ideas that we've learned from and been inspired by in 2021:
Improving Social Media: Content Moderation & Democracy
with Sarah T. Roberts (Co-founder, Center for Critical Internet Inquiry, Associate Professor of Information Studies, UCLA) and Murtaza Shaikh (Policy Lead, Online Hate, Terrorism, & Incitement at Ofcom)
David Polgar spoke with Sarah T. Roberts (Co-founder, Center for Critical Internet Inquiry, Associate Professor of Information Studies, UCLA) and Murtaza Shaikh (Policy Lead, Online Hate, Terrorism, & Incitement at Ofcom) on January 21, 2021 for Improving Social Media: Content Moderation & Democracy, part of our livestream series with TheBridge with follow-up resources and a podcast provided by The Radical AI Podcast.
Sarah outlined the concept of online commercial content moderation, referencing the organizing principle of her book, Behind the Screen: Content Moderation in the Shadows of Social Media, to describe how its purpose is fundamentally about “brand management and liability mitigation for platforms that are maintaining a business to business relationship with their actual clients and customers who are other corporate entities, advertisers, data miners, etc.”
She described how, in service to the goal of creating and keeping engagement, these social media platforms are not mirrors reflecting society’s values and interests but are instead magnifying glasses, algorithmically amplifying the most polarizing and engaging content, giving artificial prominence to particularly damaging facet of society and metaphorically angling and directing that heat in ways that ultimately starts socially destructive fires – “and we're the insects that are getting burned.”
She spoke about the rise of social media platforms as a quasi-public square or public sphere in contrast to the decline of public spaces and public institutions like public school systems and public libraries in the U.S., as a way to be engaged, to debate, to be informed, etc.
She laid out the global impact of Section 230 and what it means for social media companies and commercial content moderation that it allows firms to “dial in their own preferred tension level” where they can have no liability whether they choose to act and remove content or not, “with with a very small carve out…for some of the most heinous kind of material that's illegal everywhere in the world anyway.” She described how these firms are including similar language in their international trade agreements, which are kept private with no public input.
She described the global soft power of social media companies that invent their own judiciaries, and how these private companies are in effect taking “American jurisprudence, American norms, American sensibilities, and American legal traditions and interpretations” – specifically as interpreted by the business interests of silicon valley – “and applying them carte blanche around the world in ways that might not fit or that might actually contradict the norms, processes, expectations, and laws in other parts of the world, which have their own sovereign rights.”
Murtaza emphasized the need for social media companies to enforce their own existing content policies, lamenting that, because of a lack of transparency and accountability, the public doesn’t have any guarantee that they're being implemented or implemented consistently, and that this must be addressed before we can start talking about oversight mechanisms, regulatory mechanisms, and legislating to limit their power.
He described an oversight board as being like a supreme court in that it doesn't deal with every appeal but only with cases that are going to set precedence in terms of content policies. He mentioned that hate speech doesn't have an internationally agreed upon definition and that it’s commonly defined along the lines of a direct attack based on a number of protected characteristics.
“The issue isn't whether there should be an exemption from liability or not…[but] can laws be crafted specifically for social media companies to hold them to account to or to make them responsible in an appropriate way which does not engage normal criminal laws.”
Julie Inman-Grant
Australia's eSafety Commissioner on Safety by Design
Julie Inman-Grant, Australia's eSafety Commissioner, spoke with David Polgar about Safety by Design on Jul 6, 2021.
As the world's first online safety regulator, Julie’s role is to help Australian citizens have more safe and positive experiences online through services and regulatory schemes to take down seriously harmful content. Her office has the nation's hotline for child sexual abuse material, and they also deal with pro-terrorist content and in the wake of the Christchurch massacre. In addition, her office has significant ISP blocking powers in the event of an online crisis event as well as powers to take down image-based abuse known as revenge porn.
She talked about how Section 230 is outdated and suggested the need to start from scratch and look at the environment comprehensively to really tackle these issues, as “we're never going to regulate or arrest our way out of these online toxicity issues.”
She asserted that technology providers and platforms have a role to play in safety by design to mindfully assess risk at the front-end to build in the safety protections, imploring companies to “bake them in rather than bolt them on after the damage has been done.”
“If you build the digital roads, you need to be erecting the digital guard rails…policing those roads for dangerous drivers…[and] not only do the platforms provide these digital worlds but they also provide the vehicles.”
She described how companies tend to move when there's a genuine threat to their revenue to their reputation or to regulation, and that safety has always taken a backseat to security and privacy.
She talked about her background at Microsoft and then Twitter, describing implementing innovations like the mute button and enabling third-party reporting, reducing the time it takes to report harmful content from a few minutes down to 30 seconds and removing the burden of proof from the victim in cases of “revenge porn.”
She emphasized the importance of a change in culture around safety by design, changing the ethos to engaging in safety by design and doing this *for* industry rather than *to* industry, and discussed the importance of “coopetition” in cooperating with other companies on issues of societal good or around significant harms like child sexual abuse material or other volumetric cross-platform attacks like “revenge porn”.
She declared that the “hallmark of any robust approach to safety by design…is through transparency and accountability,” and that “unless companies do start to do better and lift their overall standards around online safety, they are they're going to be regulated, and they're going to be regulated in ways that they don't want to be regulated.”
Ask a Mentor with Renée Cummings
(Criminologist, AI Ethicist, Data Activist in Residence at University of Virginia)
Renée Cummings is the data activist in residence at the School of Data Science at the University of Virginia where she brings the JEDI principle to AI – justice, equity, diversity, and inclusion.
“Where there is justice, there is equity; where there is equity, there is diversity and inclusion; and where there is inclusion, we are looking at gender, we are looking at sexuality, we are looking at geography, we are looking at access, we are looking at age…”
Renée spoke about her background in criminology and her realization that society needs criminologists to be more involved in data science, and data scientists need to understand criminology when they're designing tools, in light of how we overestimate the risks of black and brown defendants and we over-police the same communities with predictive policing.
Renée described her career trajectory and extolled the value of speaking on topics she’s passionate about, like the challenges of predictive policing as well as civil liberties and data rights, which she predicts is probably going to be the civil rights movement of this generation.
“Data activism really is about bringing the requisite levels of advocacy and evangelism to the question of data, and understanding that we've got to challenge the data infrastructure – the power, the privilege, and the politics behind data, and how we deploy our data-driven systems in society.”
Renée spoke about the importance of ethical resilience and ethical vigilance to ensure that it’s understood that something that is as transformational as algorithmic decision-making can both save lives and improve lives but can also hurt people and destroy lives.
Ask a Mentor with Savannah Badalich
(Policy Director, Discord)
Savannah’s role as policy director at Discord involves three areas of work. First, platform policy, which basically entails questions around content policies, like defining the types of content and behaviors that are or are not allowed on the platform. These questions deal with harassment, extremism, hate speech, cyber crime policies, and other issues, and Savannah’s team updates and adjusts these policies as needed, especially based off user feedback.
The second area is product policy. Product policy involves thinking about how ideally and responsibly the product should behave and doing risk assessments with product managers as they're designing the product to mitigate potential harassment or abuse factors from the very beginning – safety by design.
The third area is public policy. Savannah works with governments, regulators, and public officials to make sure that Discord is responding to various regional concerns.
Savannah discussed the responsibility – and the significant amount of power – that private companies like Discord have in cultivating, curating, and positioning speech as well as determining which users might be de-platformed.
She acknowledged that policies are not necessarily one size fits all, and that depending on context, the same speech can communicate different things. For instance, a slur in one context could be benign in another context.
Savannah also discussed the role of pseudonymity on Discord. The ability to create personas with different names on different servers can allow users to explore different parts of their identities and find belonging within various communities. Teens especially like this feature, which enables users to explore without feeling like they'll be necessarily targeted. On the flipside are the ways in which people try to oppress others using tactics for radicalization where a sense of belonging can be harmful.
Savannah described her job search, which involved applying at dozens of companies. A pivotal moment was getting nearly to the last round during an unsuccessful bid at Twitter, and asking the recruiter for feedback. The recruiter recommended that Savannah work more with products and tech, so she worked with tech startups for a year and then reapplied, becoming a product trust partner at Twitter.
Savannah’s role at Twitter involved working with recommendation systems, drawing on her knowledge of how people can be radicalized and how dangerous narratives can be normalized through recommendation algorithms. She learned from product managers about the technical aspects of the work.
Savannah says that her biggest asset in her work at tech companies has been her earlier work community organizing and stakeholder engagement work. She had worked for a non-profit that worked on tech for the public good, Civic Hall, and she says that social services, nonprofit, and direct service work is the most needed skill set for tech right now – not just for policy or for trust and safety, but also for product managers and engineers.
“There are many different approaches to advocacy. I've been doing it mostly outside these organizations, trying to ask them to change, and this time I wanted to see what was it like to be an advocate within the system…If I can ask questions and ask for all the background documentation, I can learn enough to have the conversations.”
Savannah discussed having policy-aligned product builds that bring in trust & safety conversations early in the product development process, during the ideation sessions and the design sprints, instead of trust & safety and policy being a “janitor” cleaning up issues after the fact.
Pearlé Nwaezeigwe
(TikTok, Product Policy Manager SSA)
Pearlé Nwaezeigwe (TikTok, Product Policy Manager SSA) shared her Responsible Tech Journey on Sep 15, 2021 during All Tech Is Human’s Responsible Tech Summit.
Pearlé studied law, intending to be a human rights lawyer with ambitions of working at the United Nations, until one of her grad school professors at UC Berkeley sparked her interest in the responsible tech space, and she turned her attention to the intersection of human rights and technology. Equipped with her legal background and a burning desire to go into tech policy, she began attending as many conferences and networking events as she could. She found that tech policy wasn’t as popular as some other roles, but she kept at it and eventually met people in the responsible tech space at conferences and online. Her tireless networking provided ample opportunities to ask questions to learn about various companies’ approaches to safety issues. Fortified with this knowledge and a robust network, she began to apply to jobs. Her job search process, which took almost a year and required a lot of resilience, involved reaching out to people whenever she saw a new job, getting a referral, and then landing an interview. During interview prep, she read resources and studied the company’s community guidelines as well as critiques of their safety approaches.
Pearlé shared five key learnings:
Responsible tech is still growing – there is so much space for you to find a niche.
Your background or skills are very transferable – because nothing is set in stone, anyone with any background can land a role.
Your network is your net worth – applying online is a thing of the past. Get networking!
Self-study is key – know everything about the space you're interested in; there are resources out there for free. Read and talk about it; the more you talk about it, the more you understand it.
Don't give up – trust the process.
Arushi Saxena
(Twitter, Product Marketing Manager, Misinformation and Civic Integrity)
Arushi Saxena (Twitter, Product Marketing Manager, Misinformation and Civic Integrity) shared her Responsible Tech Journey on Sep 15, 2021 during All Tech Is Human’s Responsible Tech Summit.
In May of 2021, Arushi completed an interdisciplinary masters completely unrelated to tech. Her educational background was in finance, and she started her career working at a startup as a business operations associate that worked with brand loyalty. She utilized the double diamond approach to explore everything she could within the community, taking this cross-functional approach in order to explore the topic from different perspectives. She launched ekminuteproject.com, having experienced the consequences of misinformation in her personal life as a South-Asian and applying her financial expertise to the misinformation context. Arushi then brought that approach to Twitter. For job seekers, Arushi recommended personalizing and customizing one’s pitch to various recruiters and hiring managers, noting that her misinformation pitch was different from her private sector experience pitch. She advocated for those interested in entering the field of responsible tech to embrace the fact that we can choose how to build our stories and apply it to the tech context. Because there is no single prescribed background requirement for entry into the field, we can determine the ways in which we add value to tech.
Improving Digital Spaces panel
with Lisa Roman (VP Public Policy & Social Impact at Bumble), Lybra S. Clemons (Chief Diversity Officer at Twilio), Acya Ariyoruk (Director of Global Partnerships and Communication at Soliya, moderated by Rey Allie (Head of Trust & Safety at IRL)
The Improving Digital Spaces panel with Lisa Roman (VP Public Policy & Social Impact at Bumble), Lybra S. Clemons (Chief Diversity Officer at Twilio), Acya Ariyoruk (Director of Global Partnerships and Communication at Soliya, and moderator Rey Allie (Head of Trust & Safety at IRL) took place on Sep 15, 2021 as part of All Tech Is Human’s Responsible Tech Summit.
Lisa Roman shared how the founder of Bumble recognized the need for a dating app centered around women, with security at its core. She described the reality that companies are free to determine what they will and will not tolerate on their platforms, and that Bumble designs public policy to address specific harms, such as body shaming. She described how Bumble’s goal is to create a platform that enables users to interact in real life, so this fact is taken into account in the product design.
Lybra S. Clemons described Twilio as a developer platform, where developers are initiating the conversations. Lybra’s goal is to ensure that, each day, Twilio is becoming the most anti-racist organization that it can be. Twilio emphasizes user safety and privacy, and works to build trusted communications. Lybra described a real concern around ensuring that what users are sharing is true and ensuring that people can trust the information they are receiving. Lybra outlined her role at Twilio, describing intentionally trying to integrate anti-racist principles into everything that the company does, as well as describing the challenges of actively trying to promote equity to become an anti-racist organization when people tend to react aggressively to the realization that we live in a society that was designed to be racist. She emphasized her efforts at the company to shift DEI to be about more than hiring two women and one gay person, for example, but to embrace a systemic change.
Ayca Ariouruk described the goals of Soliya, an NGO founded almost 20 years ago after 9/11 that encourages youth from opposing political camps to meet and engage in critical thinking and talk about issues. The organization tries to convince educational professionals that technology can be helpful in scaling this approach. Ayca emphasized the need for creating a trustworthy space where people are not walking on eggshells. Soliya uses the online world as a learning tool and considers the tools and features that people can tap into online that would not be possible otherwise, including the importance of anonymity in enabling students to be transparent and to ask questions in an honest manner that they would normally not feel comfortable asking. For this reason, privacy is a core value for Soliya. With participants from authoritarian countries, Ayca emphasizes the need to make sure that people can be open about their views, trusting that their safety is protected. Ayca hypothesizes that quick workshops and lectures won’t work well, that people need to have connections, and that human contact if facilitated and done correctly can help overcome people’s prejudice.
Patricia Cartes Andres
(Head of Trust & Safety at VSCO)
Patricia Cartes Andres, Head of Trust & Safety at VSCO, shared her Responsible Tech Journey on Dec 1, 2021 as part of All Tech Is Human’s Responsible Tech University Summit.
She studied translation and interpreting in college and joined Google in Ireland at the age of 21. In 2006, prior to widespread user-generated content, online abuse was confined to manipulation of search engines like spam and keyword stuffing. By 2008, Facebook was establishing headquarters in Europe and called Patricia to join their site integrity team. She was one of two employees in Europe tackling trust and safety, which was called user operations at the time. Patricia described giving people the ability to report content for the first time, without official community standards. Patricia described the casual process as herself and a colleague looking at one another and asking, “do you think we should take this down?” Over time, they developed public policy efforts and communications efforts, engaging with regulators and law enforcement and court authorities reaching out about pieces of content. She joined the public policy team before moving on to Twitter to do similar work setting up some of this support infrastructure and determining what the right policies should be and how best to comply with local regulation. She discussed the challenge of having tech companies based in the U.S. where free speech is such a protected element which sometimes clashes with the rest of the world, and described her work mitigating the downsides of tech companies by developing the best community standards and policy rules while trying to amplify the positive side by empowering nonprofits and civil society to harness the power of social media to bring people together and to bring their missions forward.
Institute for Rebooting Social Media:
Hilary Ross (Senior Program Manager, Harvard Berkman Klein Center)
Hilary Ross, Senior Program Manager at Harvard’s Berkman Klein Center for Internet and Society, spoke about BKC’s Institute for Rebooting Social Media on Dec 1, 2021 as part of All Tech Is Human’s Responsible Tech University Summit.
Hilary’s interest in issues around how media can support democracy and the connections between media and access to information led her to study technology policy. The Rebooting Social Media project unites law and computer science in a 3 year initiative to accelerate progress in understanding and tackling social media’s emerging problems, like a decrease in privacy, online harassment, widespread hate speech and connections to real-world violence, and problems of misinformation in scale and the impacts on democracy. These problems reach across nations and have intersectional components – media, democracy, technology.
The Institute asks questions like what specifically is broken? What are the harms? What are the metrics to measure? What are new institutions and governance models? What are the data-driven solutions? Hilary described the Institute’s work around make-a-thons; fellowships for builders, academic research, and theory; work with student researchers and visiting scholars.
Responsible Tech Practitioners panel
with Dr. Ellie Sakhaee (Sr. Program Manager, Responsible AI at Microsoft), Danise Olague (Sr. Program Manager, Accessibility & Product Inclusion at Yahoo), Aimee Louise Bataclan (Communications Manager at Partnership on AI), with moderator Veronica Irwin (San Francisco Examiner Reporter)
The Responsible Tech Practitioners panel with Dr. Ellie Sakhaee (Sr. Program Manager, Responsible AI at Microsoft), Danise Olague (Sr. Program Manager, Accessibility & Product Inclusion at Yahoo), Aimee Louise Bataclan (Communications Manager at Partnership on AI), with moderator Veronica Irwin (San Francisco Examiner Reporter) took place on Dec 1, 2021 as part of All Tech Is Human’s Responsible Tech University Summit.
Dr. Sakhaee gave advice on breaking into the field of Responsible Tech, encouraging aspiring responsible technologists to start with learning some basics on AI/ML through accessible educational resources like Coursera or other MOOCs. With her own tech background, Dr. Sakhaee had to quickly learn the ins and outs of policy to round out her knowledge base. She recommends doing a lot of research, deeply but narrowly, like working exclusively on computer vision in autonomous driving, for example. She worked at TechCongress prior to her role at Microsoft, and her advice on choosing a tech company was to remember that it can be hard to judge exactly what’s going on inside a company behind closed doors; even talking to employees still offers only a limited sample set. She recommends instead asking “Which companies are open to change?” rather than “Which companies have a good ethical culture?”, and posits that it may be easier to have an impact within companies that are open to change. She offers the caveat that even within a company, interest in responsible tech can vary from team to team, and suggests that raising a narrowly scoped flag can still have a positive impact that can be replicated through an organization.
Danise emphasized the importance of critical thinking. Her background in Ethnic Studies enabled cross disciplinary analysis on race, society and oppression, prompting her to ask questions like, “How can we innovate?” and “How can we be creative to solve problems?” She had the opportunity to pivot from the field of teaching into EdTech, after interning at Common Sense Media while pursuing her MBA, and she endeavored to be a part of making the tech products herself instead of scrambling to adjust once the products had already been built. Danise had no technical background and taught herself UX design and how to build products. For her, there wasn’t one lightbulb moment, but lots of years of experiences that developed her understanding of how close she wanted to be to the impact.
Aimee previously worked on PR in tech companies as well as working in the climate space. PAI works with a lot of different tech companies, across all levels of maturity in their experiences with and approaches to ethical tech practices. She suggests that one’s ability to make an impact within a tech company comes down to how much that company is open to change and how much a particular individual might be able to affect that change. Her advice to those working within a company or organization who see an opportunity for change is to build alliances within the company. She offers that “you are likely not the only one,” and finding other people who are also interested in the same thing can enable people to work together to affect change.
