Daily Archives: April 6, 2021

There are a few basic rules of working in news. First tell the truth and tell it well. Second always keep your eyes and ears peeled for new stories. Lastly, but certainly not least keep a pulse on old stories for new developments. I think my team at ABC does a great job of all three especially keeping up with the stories that weve taken pride in telling over the years.

Thats why it was pretty surprising when an article called The New Menendez Defenders showed up in the New York Times in February.

There were no new developments in this case - not in years. But this news wasnt about the infamous Menendez brothers themselves, or their parents, or the trial that created a media frenzy decades ago. This story was about the army of internet defenders the convicted murderers had garnered on wait for it TikTok.

Lyle Menendez, left confers with brother Erik during a court appearance, April 2, 1991, in Beverly Hills, Calif.

In order to understand where weve arrived now, you have to understand the basic facts of the Menendez Brothers case. Lyle and Erik Menendez grew up in Southern California as sons of a wealthy couple, Kitty and Jose Menendez. In 1996, a jury found the brothers who were in their early 20s - guilty of first-degree murder of their parents.

Watch the full story on "20/20" FRIDAY at 9 p.m. ET on ABC.

Both Erik and Lyle say they committed the murders after years of insufferable sexual abuse at the hands of their father. Meanwhile, many newspaper headlines at the time and documentary programs in the years following capitalized on the concept of greedy rich kids killing their parents for money.

The brothers have now served 31 years in prison, most of that time spent apart from each other - only being reunited recently at a prison in San Diego. Theyve exhausted most appeals processes and their prospects of a new trial have dwindled. For them, life inside prison walls hasnt changed all too much. But out in the real world, where social media sites like TikTok appear to have garnered the power to create actual change in the social justice system, a lot has changed.

The New York Times article detailed the mass following and attention the Menendez Brothers had gained on TikTok. As a millennial who only uses TikTok to watch videos of cute dogs and easy recipes, I was caught off guard. Why the Menendez Brothers? Why now? What is there even to post about?

I took to the primary source TikTok itself. I typed Menendez Brothers into the search bar, and my screen flooded with hundreds of videos. But the ones that I first saw didnt have anything to do with the Menendez Brothers case at all. Instead, a young-looking girl was on my screen, fawning over images of the young, attractive brothers at their murder trial.

Sultry music played behind slowed-down footage as she joked about them being the hottest criminals ever. Most of the videos I clicked on followed this theme, and I began to wonder if this whole New Age Menendez Defender movement was nothing more than teenage girls with an extreme attraction to bad boys. But then the TikTok algorithm kicked in.

As you interact with TikTok videos, the app will show you more videos related to that topic. Slowly but surely, the tone of the videos showing up completely changed. Sad music played behind clips of the brothers testifying about the abuse they claim to have endured; their faces contorted in anguish.

Lyle, left, and Erik Menendez leave the courtroom in Santa Monica, Calif., Aug. 6, 1990, after a judge ruled that conversations between the two brothers and their psychologist after their parents were slain are not privileged and can be used as evidence in their murder case.

Captions begged viewers to listen to their heart wrenching testimony. Comment sections were flooded with outrage over their conviction and links to learn more information about the case. There were 173 million views on these videos. Right before my eyes, the army behind the Menendez Brothers grew stronger and louder with each passing second. But who were these people?

Frankly, they appear to overwhelmingly be kids. In researching this story, I reached out to dozens of TikTok users about the videos they were posting and I could hardly find anyone over the age of eighteen. Could this be, in part, simply due to the audience that TikTok usually attracts? Of course, yes.

However, speaking to the teenagers I reached didnt feel like speaking to a teenager. Theyd greet me with a bubbly introduction and tell me how old they were, before getting into an evidence-based narrative about why they believed the Menendez Brothers sentencing was unjust. The demeanor these teenagers spoke with and the sheer amount of knowledge they had accumulated about a legally complex case was astounding.

Their youth only peeked through in their giddiness and excitement to help with my reporting. One TikTok user, eager to point me in the right direction, put me in a group chat with about ten other TikTok users who primarily used the social media site to create videos about the Menendez brothers or to connect with other supporters. I was soon flooded with messages from teens aged 13-19, complete with paragraphs-long sentiments about their beliefs. Most of it included complex legal jargon.

One of those very eloquent teenagers was an 18-year-old girl named Janne from Germany. She told me she first came across the case from a recommended video on YouTube that piqued her interest.

Janne discovered that the entire trial was available to watch online and spent a great deal of time watching testimonies and forming her opinions about the Menendez Brothers. Through her research, she says she has become an educated and passionate supporter of Erik and Lyle Menendez taking to TikTok to share her opinions.

Janne from Germany said she first came across the Menendez case from a recommended video on YouTube.

Jannes TikTok account has garnered nearly 45,000 followers. Her most popular video, which has more than a million likes, shows the brothers testifying about sexual abuse while a sad song plays in the background. Janne does not credit the success of her account to advanced video editing skills or social media strategy but rather to the sheer shocking and upsetting nature of Erik and Lyles testimony.

Watching someone describe these types of experiences is always very sad and very compelling, she told me.

That has been one of the most interesting parts about this reporting. None of these teenagers deny that this horrific crime of their parents being murdered happened or that it was Erik and Lyle who committed it. Rather they take an empathetic approach supported guided by the benefits of hindsight.

As New York Times reporter Ezra Marcus told me these young adults werent around to experience the sensationalism that takes place when a crime initially happened. They didnt live through the shock of the crime itself, or even the news reports in the days and weeks to follow. Theyre now looking at the murders through the lense of the values of 2021 especially with regard to the claims of sexual abuse.

All of the TikTokers I spoke with emphasized that they have grown up in an era where #MeToo reigns and where people can speak more openly about what were previously deemed to be uncomfortable issues.

Lyle Menendez, right, and brother Erik listen to a charge of murder conspiracy against them, Dec. 29, 1992, in Los Angeles.

These realities, combined with the abundant amount of available courtroom footage and unique style of TikTok have laid the groundwork for Gen Z TikTok users to generate an entirely new category of content.

Gone are the days where a news station was the only place to get information. Now, these TikTok users are thriving on the ability to tell a brand new story from a new perspective using old footage.

In a matter of months, the Menendez Brothers went from being an old tale of true crime to a viral social media sensation. Todays TikTok activists may be teenagers, but they believe they know what theyre talking aboutand a few million likes may prove it.

Visit link:

The Menendez brothers are back in the zeitgeist thanks to teens on TikTok: Reporters Notebook - ABC News

It is typical of our faculty to come up with something as bold and innovative as this," said then-MIT president Charles Vest at a special gathering of community members and press in April 2001. OpenCourseWare looks counterintuitive in a market-driven world. It goes against the grain of current material values. But it really is consistent with what I believe is the best about MIT ... It expresses our belief in the way education can be advanced by constantly widening access to information and by inspiring others to participate."

In the 20 years since it began, MIT OpenCourseWare has become a pillar of the open education community, an exemplar of the MIT ethos, and an invaluable resource to millions of learners around the world. People of all ages and all walks of life have used the lectures, videos, problem sets, and other content to pursue their curiosity and passions, improve their careers, and get a leg up in their studies. Now, the team looks to the future with a clear sense of purpose, informed by the learning needs underscored by the Covid-19 pandemic.

OpenCourseWare launched during the early days of Web 2.0 and a growing but highly commercialized interest in e-learning. Charles Vest had commissioned the Lifelong Learning Committee, asking its members to propose an educational technology project that would extend MIT's reach beyond classrooms. That committees recommendation was to launch OpenCourseWare, a website offering all of MITs course materials, available for free to anyone. Within one year, OCW had published a pilot website with 50 courses and attracted worldwide acclaim. Today, OCW offers materials from over 2,570 courses spanning the MIT graduate and undergraduate curriculum, from 1,735 MIT faculty and lecturers from 33 academic units across all five schools, including syllabi, lecture notes, problem sets, assignments, and audiovisual content including recorded lectures. To date, OCW has been a resource for over 210 million unique users, with over 70 percent of users in 2020 coming from outside the United States.

Professor Dick KP Yue, who chaired the Lifelong Learning Committee, described the impetus for the project in the proposal: "In the digital age, institutions like MIT have a responsibility and an opportunity to impact learners far beyond their campuses. OCW embodies MIT's commitment to constantly widening access to knowledge."

The value of that commitment is borne out by learners who have shared their stories over the years from Tooba Siddiqui in Pakistan, who had access to education through OCW when other doors were closed to her, to Anita Moreno in Nevada, who used OCW to keep up with her studies following a brain aneurysm. I cannot emphasize enough how this site has boosted my confidence, that I am still able to comprehend and succeed in an engineering program, says Moreno.

Today, Professor Krishna Rajagopal, dean for digital learning, says, It was the best thing MIT could have done at that moment for MIT and for the world."

A revolution in the making

From its modest and experimental beginning, OCW sparked a new era in the growing open-education movement. Beyond the courses themselves, OCW has had a broad impact on the way online learning resources have evolved in higher education, setting the template for other colleges and universities undertaking similar efforts and helping launch the open education resources (OER) revolution.

Free access to knowledge is a powerful foundation for progress, says OCW Director Curt Newton, but its not the whole picture. OER that lifts up everyones right to contribute to shared knowledge, and builds everyones capacity to extend that knowledge, is creating new paths for us to work together on the worlds most important, complex, and rapidly evolving challenges.

It helps that OCW and Creative Commons share family bonds. Launched the same year, electrical engineering and computer science Professor Hal Abelson was a member of the founding teams for both, and helped arrange for OCW to be the first institutional project to use Creative Commons licenses, In turn, OCWs early adoption of Creative Commons licenses helped demonstrate their usefulness and lent credence to the burgeoning open movement.

In 2005, OCW helped launch the OpenCourseWare Consortium (now Open Education Global), whose network of over 300 higher education institutions and related organizations have freely shared many thousands of courses, open textbooks, and other resources, and collaborated to foster widespread adoption of OERs.

At MIT, OCW has paved the way for other innovative new learning platforms, such as MITx and MicroMasters, Open Learning Library, and professional and executive learning programs.

Integrating teachers into the experience of OCW was a key priority very early on. (Indeed, the faculty committee originally envisioned OCW as being used by educators almost exclusively its widespread popularity among students and lifelong learners was a welcome surprise.) Educators around the world have shared their experiences of using OCW to master new content or inspire and engage students. In 2013, the team launched the OCW Educator program; now, hundreds of OCW courses include Instructor Insights sections where faculty share how they have taught their courses through text, video, and most recently, through the Chalk Radio podcast.

A core tenet of MITs mission is to create and share knowledge, empowering our own community and myriad others to bring this knowledge to bear on the worlds great challenges. From its inception, OpenCourseWare has offered a new and substantial way of realizing that mission in the 21st century, says Rajagopal. For individual learners, OCW is a means to expand understanding and satisfy curiosity, to support personal and professional goals, or change in their communities; for educators, it's a resource library to help augment and strengthen their curricula, enriching the experience of so many students; for educational organizations, its an invitation to nurture a shared commitment to open knowledge. OCW provides invaluable resources for millions and paves the way for others to contribute in their own ways to sharing and using knowledge for the betterment of humankind.

Living legacy

The courses on OCW have also come to reflect the way that MIT, and its relationship with the world, has grown and changed over the last two decades. Perhaps its no surprise that 6.0001 (Introduction to Computer Science and Programming in Python) and 18.06 (Linear Algebra) are consistently among the most-viewed courses. But the ebb and flow of traffic on OpenCourseWare reflects topics in the zeitgeist, too. When Esther Duflo and Abhijit Banerjee won the Nobel Prize for Economics in 2019, Duflos OCW course The Challenge of World Poverty spiked in popularity as people all over the world sought to understand her worldview and learn from her. A similar spike occurred with Professor Gary Genslers graduate-level course Blockchain and Money when he was nominated to serve in the Biden administration earlier this year.

As MIT faculty have endeavored to offer students a nuanced understanding of our complex world, so too have the materials on OCW grown more expansive. Learners around the world can now delve into Ethics for Engineers: Artificial Intelligence, Queer Cinema and Visual Culture, and Black Matters: Introduction to Black History alongside engineering, math, and science standbys.

As instructors, were excited about what we get to do in the classroom with our students, and also its wonderful to have the opportunity to expand beyond the classroom, and to actually make the material that we develop for students available more broadly, says Amah Edoh, assistant professor of anthropology and African studies and a member of the OCW Faculty Advisory Committee, whose courses Africa and the Politics of Knowledge and Global Africa: Creative Cultures are on OCW.

She also appreciates the role OCW plays in sharing MIT expertise with the world beyond engineering and science. My particular investment in this has been around making African studies more visible at MIT ... It really changes the image, the idea that others have of MIT. I think its particularly important to show that we also have courses in African studies specifically. Oftentimes Africa, African countries, African people are seen not as agents, but rather as a space where you go to solve problems. So to show that we can engage knowledge production in and on Africa critically, to me is very important.

In capturing course materials and videos, OCW does more than open windows into MIT for global learners; it also provides a unique, living archive of teaching at MIT. Alumni can revisit favorite classes or share them with colleagues, peers, or kids looking into college. High-schoolers can get a sense of what courses in different fields will entail. Faculty whose course materials are preserved on OCW have an artifact of their teaching legacy, for everyone, forever. Consider How to Speak, a OCW video of the late Professor Patrick Winstons beloved Independent Activities Period course, which he taught for 40 years before passing away in 2019; posted in December 2019, it has now been viewed 3.3 million times.

Looking forward

OpenCourseWare enters its third decade on the heels of unprecedented global disruption. During the first months of the Covid-19 pandemic, when schools and businesses closed and billions of people around the world sheltered in place at home, traffic to OCW spiked to 2.2 million visits a month, a 75 percent increase from 2019. Since then, site visits have settled into a 15 percent uptick in use. More importantly, the massive shift to remote and hybrid learning over the past year has brought into sharp relief both the opportunities of online education and the disparities of access, technology, and equity for learners everywhere. In charting a course for the future, the OCW team has the opportunity to draw on 20 years of experience in addressing the issues brought to the fore in 2020.

The first years of OCW have been primarily about the power of access, says Newton. A core principle of where were heading in our upcoming program is the progression from giving access to knowledge to really driving towards educational equity.

Later this year, OpenCourseWare will launch its NextGen platform and program. Its three principal aims are offering a vibrant reflection of MIT education as it evolves, delivering a more user-focused design and experience, and broadening access and usability to a larger global population. The NextGen platform will support a more dynamic experience of OCWs robust multimedia content, allowing users to seamlessly search, browse, download, remix, and redistribute all materials more easily. Individuals can get a sneak peek of the new OCW and sign up to be a beta tester.

Another major pillar of the NextGen platform is mobile optimization, a user-friendly interface to provide readable, searchable content on any device. With 92.6 percent of internet users around the world using mobile devices at least some of the time, and with smartphone use growing at a rate of 7 percent per year, this change represents not only a catch-up to current need but also a purposeful approach to finding and engaging with future learners.

As we look at the next year, five years, 20 years of OpenCourseWare, our goal is to keep pace with the evolving artifacts of MIT teaching and learning, offering the best possible experience to our growing community of learners, says Newton. We are also committed to continually reinvesting in the OER community working collaboratively to share resources and engage with the people and organizations at the vanguard of access and equity in education.

MIT Open Learning will host an online celebration of OpenCourseWares 20th anniversary on Wednesday, April 7, from noon to 1 p.m. EDT featuring OCW leadership, MIT faculty, and learners sharing stories and ideas about the past, present, and future of open education, at MIT and beyond.

Read the original:

How MIT OpenCourseWare became an educational resource to millions around the world - MIT News

That show debuted in 1999. But it was around the turn of the new century, and concurrent with the rise of Silicon Valley fortunes, when a few zeros got added in the popular imagination. Billion replaced million as the impossibly large yet routinely cited number.

Imagine bestowing a TV show with the title Billions in 1986 or 1996 it would have seemed hubristic if not wholly fictional. By 2016, when Billions debuted on Showtime, it captured the zeitgeist. Forbes called the 2010s A Decade of Billionaires, and indeed, so many were minted in those years that the count nearly doubled between 2008 and 2020. There are 2,095 billionaires today, according to the publication.

Unless you were a Wall Street quant, however, trillion had long felt remote. It was like gazillion: a joke number. The Trillionaire Next Door is what Andy Borowitz, the humorist, titled his spoof 2000 book about day trading. In a radio appearance at the time, Mr. Borowitz said he wrote it because, People aim too low. I mean, a millionaire? Please.

That same year 2000 the word trillion appeared in The New York Times 856 times.

In the first three months of 2021, trillion has already appeared in this paper 723 times.

Much of that usage is in reference to the $1.9 trillion stimulus bill that Congress passed in March. (Too much? Too little? The debate goes on). Faster than you could say, er, 1 followed by 12 zeros, the Biden administration had already turned its attention to an infrastructure bill it hopes to pass later this year. The amount? $2.2 trillion.

Indeed, its not just money. The United Nations Environmental Programs worldwide tree-planting movement, nicknamed the billion tree campaign when it started in 2006, has become the trillion-tree campaign. Some artificial intelligence executives are planning for a future of trillions of internet-connected devices. And scientists are already talking about Brood X, a colony of cicadas that will emerge from their 17-year hibernation to flood the United States this spring.

Whats the high end for the number of cicadas we can expect? One guess.

See original here:

How Many Is a Trillion? - The New York Times

Conversations around AI and ethics may have started as a preoccupation of activists and academics, but now prompted by the increasing frequency of headlines of biased algorithms, black box models, and privacy violations boards, C-suites, and data and AI leaders have realized its an issue for which they need a strategic approach.

A solution is hiding in plain sight. Other industries have already found ways to deal with complex ethical quandaries quickly, effectively, and in a way that can be easily replicated. Instead of trying to reinvent this process, companies need to adopt and customize one of health cares greatest inventions: the Institutional Review Board, or IRB.

Most discussions of AI ethics follow the same flawed formula,consisting of three moves, each of which is problematic from the perspective of an organization that wants to mitigate the ethical risks associated with AI.

Heres how these conversations tend to go.

First, companies move to identify AI ethics with fairness in AI, or sometimes more generally, fairness, equity, and inclusion. This certainly resonates with the zeitgeist the rise of BLM, the anti-racist movement, and corporate support for diversity and inclusion measures.

Second, they move from the language of fairness to the language of bias: biased algorithms, as popular media puts it, or biased models as engineers (more accurately) call it. The examples of (allegedly) biased models are well-know, including those from Amazon, Optum Health, and Goldman Sachs.

Finally, they look for ways to address the problem as theyve defined it. They discuss technical tools (whether open source, or sold by Big Tech or a startup) for bias identification, which standardly compare a models outputs against dozens of quantitative metrics or definitions of fairness found in the burgeoning academic research area of machine learning (ML) ethics. They may also consider engaging stakeholders, especially those that comprise historically marginalized populations.

While some recent AI ethics discussions go beyond this, many of the most prominent dont. And the most common set of actions that practitioners actually undertake flows from these three moves: Most companies adopt a risk-mitigation strategy that utilizes one of the aforementioned technical tools, if theyre doing anything at all.

All of this should keep the stewards of brand reputation up at night, because this process has barely scratched the surface of the ethical risks that AI introduces. To understand why this is, lets take each of these moves in turn.

The first move sends you off in the wrong direction, because it immediately narrows the scope. Defining AI ethics as fairness in AI is problematic for the simple reason that fairness issues are just a subset of ethical issues youve just decided to ignore giant swaths of ethical risk. Most obviously there are issues relating to privacy violations (given that most of current AI is ML, which is often powered by peoples data), and unexplainable outputs/black-box algorithms. But there are more. For example, the primary ethical risk related to AI-powered self-driving cars isnt bias or privacy, but killing and maiming. The ethical risk of facial recognition technology doesnt end when bias has been weeded out from the model (of which there are a number of examples); a non-biased facial recognition software still enables surveillance by corporations and (fascist) governments. AI also requires vast amounts of energy to power the computers that are training the algorithms, which entails a shocking degree of damage to the environment. The list of ways a company can meet with ethical disaster is endless, and so reducing AI ethics to issues involving fairness is a recipe for disaster.

The second move further reduces your remit: Issues of bias are a subset of issues of fairness. More specifically, issues of bias in the context of AI are issues of how different subpopulations are treated relative to others whether goods, services, and opportunities are distributed fairly and justly. Are the job ads placed in such a way that they are just as likely to be seen by the African-American population as by the white population? Are women applying for a job just as likely to have their resumes lead to an interview as a mans resume? The problem with this approach is that issues of fairness extend beyond issues of fair distributions of goods to various subpopulations.

Most obviously, there are issues about what any individual deserves independently of how others are treated. If Im torturing you, and you protest, it would hardly justify my actions to say, Dont worry, Im torturing other subpopulations at equal rates to the population of which youre a member. The entire category of human rights is about what every person deserves, independently of how others are being treated. Fairness crucially involves issues of individual desert, and a discussion, let alone a risk-mitigation strategy, that leaves this out is the more perilous for it.

The final trouble for organizations arrives in the third move identifying and adopting bias-mitigation strategies and technical tools. Organizations often lean on technical tools, in particular, as their go-to (or only) meaningful instrument to ferret out bias, as measured by quantitative definitions of fairness found in recent computer science literature. Here, we run into a raft of failures at ethical risk mitigation.

First, those two-dozen-plus quantitative metrics for fairness are not compatible with each other. You simply cannot be fair according to all of them at the same time. That means that an ethical judgment needs to be made: Which, if any, of these quantitative metrics of fairness are the ethical/appropriate ones to use? Instead of bringing in lawyers, political theorists, or ethicists all of whom have training in these kinds of complex ethical questions these decisions are left to data scientists and engineers. But if the experts arent in the room, you cannot expect your due diligence to have been responsibly discharged.

Second, these tools standardly only kick in well into the development lifecycle. Because they measure the output of AI models, theyre used after data sets have been chosen and models have been trained and a good deal of resources have been devoted to the product. It is then inefficient, not to mention unpopular, to go back to the drawing board if a bias problem is detected that cannot be solved in a fairly straightforward way.

Third, while the search for a technical, quantitative solution to AI ethics is understandable, the truth is that many ethical issues are not reducible to quantitative metrics or KPIs. Surveillance is a problem because it destroys trust, causes anxiety, alters peoples behavior, and ultimately erodes autonomy. Questions about whether people are being treated respectfully, whether a product design is manipulative or merely giving reasonable incentives, whether a decision places a burden on people that is too great to reasonably expect of them, these all require qualitative assessments.

Fourth, these technical tools do not cover all types of bias. They do not, for instance, ferret out whether your search engine has labeled Black people gorillas. These are cases of bias for which no technical tool exists.

Fifth, the way these tools measure for bias are often not compatible with existing anti-discrimination law. For example, anti-discrimination law forbids companies from using variables like race and gender in their decision-making process. But what if that needs to be done in order to test their models for bias and thus influences the changes they make to the model in an effort to mitigate bias? That looks to be not only ethically permissible, but plausibly ethically required as well.

Finally, as regards engaging stakeholders, that is generally a good thing to do. However, aside from the logistical issues to which it gives rise, it does not by itself mitigate any ethical risks; it leaves them right in place, unless one knows how to think through stakeholder feedback. For instance, suppose your stakeholders are racist. Suppose the norms that are local to where you will deploy your AI encourage gender discrimination. Suppose your stakeholders disagree with each other because, in part, they have conflicting interests; stakeholders are not a monolithic group with a single perspective, after all. Stakeholder input is valuable, but you cannot programmatically derive an ethical decision from stakeholder input.

My point here isnt that technical tools and stakeholder outreach should not be avoided; they are indeed quite useful. But we need more comprehensive ways to deal with ethical risk. Ideally, this will involve building a comprehensive AI ethical risk-mitigation program that is implemented throughout an organization admittedly a heavy lift. If a company is looking for something to do in relatively short order that can have a big impact (and will later dovetail well with the bigger risk-mitigation program), they should take their cues on ethical risk mitigation from health care and create an IRB.

In the United States, IRBs in medicine were introduced to mitigate the ethical risks that arose and were commonly realized in research on human subjects. Some of that unethical conduct was particularly horrific, including the Tuskegee experiments, in which doctors refrained from treating Black men with syphilis, despite penicillin being available, so they could study the diseases unmitigated progression. More generally, the goals of an IRB include upholding the core ethical principles of respect for persons, beneficence, and justice. IRBs carry out their function by approving, denying, and suggesting changes to proposed research projects.

Comparing the kinds of ethical risks present in medicine to the kinds present in AI is useful for a number of reasons. First, in both instances there is the potential for harming individuals and groups of people (e.g. members of a particular race or gender). Second, there exists a vast array of ethical risks that can be realized in both fields, ranging from physical harm and mental distress to discriminating against protected classes, invading peoples privacy, and undermining peoples autonomy. Third, many of the ethical risks in both instances arise from the particular applications of the technology at hand.

Applied to AI, the IRB can have the capacity to systematically and exhaustively identify ethical risks across the board. Just as in medical research, an AI IRB can not only play the role of approving and rejecting various proposals, but should also make ethical risk-mitigation recommendations to researchers and product developers. Moreover, a well-constituted IRB more on this in a moment can perform the functions that the current approach cannot.

When it comes to building and maintaining an IRB, three issues loom large: membership of the board, jurisdiction, and articulating the values it will strive to achieve (or at least the nightmares it strives to avoid).

To systematically and exhaustively identify and mitigate AI ethical risks, an AI IRB requires a diverse team of experts. You will want to place an engineer that understands the technical underpinnings of the research and/or product so the committee can understand what is being done and what can be done from a technical perspective. Similarly, someone deeply familiar with product design is important. They speak the language of the product developers, understand customer journeys, and can help shape ethical risk-mitigation strategies in a way that doesnt undermine the essential functions of the products under consideration.

Youll also want to include ethics-adjacent members, like attorneys and privacy officers. Their knowledge of current and potential regulations, anti-discrimination law, and privacy practices are important places to look when vetting for ethical risks.

Insofar as the AI IRB has as its function the identification and mitigation of ethical risks, it would be wise of you to include an ethicist, e.g. someone with a Ph.D. in philosophy who specializes in ethics, or, say, someone with a masters degree in medical ethics. The ethicist isnt there to act as kind of priest with superior ethical views. Theyre there because they have training, knowledge, and experience related to understanding and spotting a vast array of ethical risks, familiarity with important concepts and distinctions that aid in clear-eyed ethical deliberation, and the skill of helping groups of people objectively assess ethical issues. Importantly, this kind of risk assessment is distinct from the risks one finds in the model risk assessments created by data scientists and engineers, which tend to focus on issues relating to accuracy and data quality.

You may also find it useful to include various subject-matter experts depending on the research or product at hand. If the product is to be deployed in universities, someone deeply familiar with their operations, goals, and constituencies should be included. If it is a product to be deployed in Japan, including an expert in Japanese culture may be important.

Lastly, as part of an effort to maintain independence and the absence of conflict of interests (e.g. members looking for approval from their bosses), having at least one member unaffiliated with your organization is important (and is, incidentally, required for medical IRBs). At the same time, all members should have a sense of the business goals and necessities.

When should an AI IRB be consulted, how much power should it have, and where should it be situated in product development? In the medical community, IRBs are consulted prior to the start of research. The reason there is obvious: The IRB is consulted when testing on human subjects will be performed, and one needs approval before that testing begins. When it comes to authority, medical IRBs are the ultimate authority. They can approve and reject proposals, as well as suggest changes to the proposal, and their decisions are final. Once an IRB has denied a proposal, another IRB cannot approve it, and the decision cant be appealed.

The same rule should apply for an AI IRB.

Even though the harm typically occurs during deployment of the AI, not research and product development, theres a strong case for having an AI IRB before research and/or product development begins. The primary reason for this is that its much easier and therefore more inexpensive and efficient to change projects and products that do not yet exist. If, for instance, you only realize a significant ethical risk from a potential or probable unintended consequence of how the product was designed, you will either have to go to market with a product you know to be ethically risky or you will have to go through the costly process of reengineering the product.

While medical IRBs are granted their authority by the law, there is at least one strong reason you should consider voluntarily granting that degree of power to an AI IRB: It is a tool by which great trust can be built with employees, clients, and consumers. That is particularly true if your organization is transparent about the operations even if not the exact decisions of the IRB. If being an ethically sound company is at the top of the pyramid of your companys values, then granting an AI IRB the independence and power to veto proposals without the possibility of an appeal (to a member of your executive team, for instance) is a good idea.

Of course, that is often (sadly) not the case. Most companies will see the AI IRB as a tool of risk mitigation, not elimination, and one should admit at least the possibility if not the probability of cases in which a company can pursue a project that is ethically risky while also highly profitable. For companies with that kind of ethical risk appetite, either an appeals process will have to be created or, if they are only minorly concerned with ethical risk mitigation, they may make the pronouncements of the board advisory instead of required. At that point, though, they should not expect the board to be particularly effective at systematically mitigating ethical risks.

Youve assembled your AI IRB and defined its jurisdiction. Now youll need to articulate the values by which it should be guided. The standard way of doing this is to articulate a set of principles and then seek to apply those principles to the case at hand. This is notoriously difficult, given the vast array of ways in which principles can be interpreted and applied; just think about the various and incompatible ways sincere politicians interpret and apply the principle of fairness.

In medical ethics and in the law, for that matter decision-making is usually not guided by principles alone. Instead, they rely on case studies and precedent, comparing any given case under investigation to previous cases that are similar. This allows your IRB to leverage the insights brought to bear on the previous case to the present case. It also increases the probability of consistency in the application of principles across cases.

Progress can be made here by articulating previous decisions senior leadership made on ethical grounds prior to the existence of the IRB. Suppose, for instance, the IRB knows senior leaders rejected a contract with a certain government due to particular ethical concerns about how the government operates generally or how they anticipated that government would use their product. The reasoning that led to the decision can reveal how future cases ought to be decided. In the event that no such cases exist and/or no such cases have been disclosed, it can be useful to entertain fictional examples, preferably ones that are not unlikely to become real examples in the future, and for the IRB to deliberate and decide on those cases. Doing that will ensure readiness for the real case when it arrives on their doorstep. It also encourages the cool objectivity with which fictional cases can be considered when no money is on the line, for instance to transfer to the real cases to which theyll be compared.

***

We all know that adopting an AI strategy is becoming a necessity to stay competitive. In a remarkable bit of good news, board members and data leaders see AI ethical risk mitigation as an essential component of that strategy. But current approaches are grossly inadequate and many leaders are uncertain how to develop this part of their strategy.

In the absence of the ideal a widespread commitment to creating a robust AI ethical-risk program from day one building, maintaining, and empowering an AI IRB can serve as a strong foundation for achieving that ideal. It can be created in relatively short order, it can be piloted fairly easily, it can be built on and expanded to cover all product teams and even all departments, and it creates and communicates a culture of ethics. Thats a powerful punch not only for AI ethics, but for the ethics of the organization more generally.

Continue reading here:

If Your Company Uses AI, It Needs an Institutional Review Board hbr.org - Harvard Business Review

The natural and organic movement has been the nexus for many elevated claims, but most notably the intersection of human health and environmental impact. The data around the human health component is becoming higher resolution and more widespread. And most of the New Hope readership is probably already of the mindset that making better choices about what we eat and wear can scale up environmental impact to save the planet. But how exactly does that happen, to what degree is it occurring and is it being optimized?

Civilization, at least as we know it, is running out of time; the United Nations stated in 2014 that at current rates of soil degradation and erosion there are only 60 harvests left. Humans have an affinity for procrastination; we are in essence facing the biggest exam of our existence. The test is tomorrow morning and we have no other choice than to start cramming right now in order to get a passing grade.

Related: 5 ways brands can tell their regenerative stories

Simultaneously, natural and organic categories are reaching positive tipping points, becoming more mainstream all the time. These new phases of growth, wider distribution options and increased consumer engagement usher in new paradigms and opportunities. When it comes to how environmental impact and soil stewardship are part of the solution, up until now, most of this discussion has been based on general assumptions or low-resolution modeling. There is an opportunity at hand to change thatto scale the measuring of environmental impact at the farm level. The old adage of what gets measured gets managed is just as true here as it has ever been. With measurement we can optimize and that makes our movement incredibly well-positioned for the future.

This new chapter that we are embarking upon is most well-known by the name regenerative agriculture. Regenerative is a synonym for net positive. So, when were talking about regenerative farming or ranching, were talking about mimicking nature to steward positive environmental outcomes while still producing the high-quality food and fiber that society needs.

Warning: The next two paragraphs get borderline nerdy for just a minute. Buckle up, well be fine!

There are many environmental benefits from regenerative agriculture, but from a climate change perspective specifically, regenerative is based on the notion of drawdown. At the simplest level, there is too much carbon dioxide (CO2) in the atmosphere. The gas traps heat and is one of the greenhouse gases (GHGs) causing climate change. The part that often doesnt make it into the broader zeitgeist is that we already have 409 ppm of CO2 in the atmosphere and scientists widely agree we need to be below 350 ppm; some say as low as 280 ppm. So, if we stop burning all fossil fuels today, were still totally screwed, climate change is already baked in. If you, like me, just hearkened back to an image of Al Gore on his scissor-lift, back in 2006 when CO2 levels were ~380 ppm, guess whatyoure on an even sicker planet than we were back then and were probably getting old. Sorry, who invited the bad news fairy anyways?

The key takeaway is that we need to stop burning fossil fuels and we need somewhere else to put these legacy carbon compounds already in the atmosphere. Incredible amounts of money are being spent on carbon capture technologies that can essentially vacuum the carbon out of the atmosphere. However, these are new and unproven technologies with many barriers to scale. Nature already has a scalable mechanism for this process; its called photosynthesis. That equation we all learned in high school biology, that no one can ever seem to remember, is probably the most important thing to the future of our existence. Plants, through the power of sunlight, are able to take carbon dioxide out of the atmosphere, combine it with water (H2O) in the soil and make carbo-hydrates. These carbohydrates form the basis of all soil organic matter. There is a whole economy of carbon trading that scientists are just beginning to understand and these organic matter compounds are the currencies. These carbohydrate-based particles are also your smallest structures in the soil and thus have the most surface area to hold on to more water. When you have more water available in an ecosystem, it can support more biodiversity.

Bare ground or exposed soil does not support photosynthesis. There are no plants present to capture the suns energy or roots to cycle minerals below ground, and the soil is left prone to wind and water erosion. Eventually, bare ground crusts over, which drastically decreases its ability to infiltrate water, and there is no supporting habitat for biodiversity. Instead, the thermal mass of the soil holds onto even more heat from the sun, further contributing to warming. Every time you see bare soil you are witnessing a missed opportunity to use photosynthesis to help reverse climate change.

Almost every brand we talk to at Land to Market starts the conversation by wanting to know the list of regenerative practices. Hopefully this doesnt burst any bubbles, but there arent practices that are inherently regenerative or those that arent. Regenerative is an outcome. It turns out, the nuance of which practices to implement on which properties mattersa lot! Its part of human nature to want formulas, panaceas, recipes, silver bullets, easy buttons, a laundry-folding machine (or maybe that one is just me). The reality is that it just isnt that simple.

Managing land to achieve net positive results is a bit like managing a business to become profitable. In fact, those who come from a triple-bottom-line perspective would point out that regenerative agriculture is actually talking about environmental profit. Every business uses different formulas and recipes, but good businesses use KPIs to track whether the formula or recipe is right for their business. Youve never heard businesses touting the profitability practices that they implement. Like businesses, agriculturalists operate on principles, things such as preventing bare soil. These principles act as a guide, but there are no silver bullets that say how exactly the farm should prevent exposed ground on their operation. Understanding and actually embracing the nuance of the specific context is critical to filtering what strategies to try first. A practice will certainly not produce the same outcome in different environments across the country, and some may not even apply across fence lines between neighbors. Just like with a business, you have to: proactively plan, set up feedback loops and KPIs, adjust accordingly and continue to optimize.

Unfortunately, we live in the most ecologically illiterate society of all time. There is still so much to learn and even relearn. When it comes to understanding the context and embracing the complexity of a piece of land, who better to do that than the farmer? They know their land and theyre frustrated by the idea of folks in far-away places telling them what they should be doing differently. The world is waking up to the reality that healthy soil can be part of the solution to so many of the worlds greatest challenges: climate change, world hunger, water insecurity and struggling rural economies. But for all of human history, farmers have been generally seen as less intelligent and relegated to the peasant class. We will not solve this problem until we change that perception. We need to collectively ask ourselves where we can empower and support and where we can get out of the way.

We approach that in multiple ways at Land to Market. We have a scientific protocol, Ecological Outcome Verification, that measures soil health, sequestered carbon, water and biodiversity the same way all over the globe. The data set is broken up into leading and lagging indicators. The leading indicators allow farmers to fine tune their management and optimize outcomes. Those that achieve net-positive results are connected with progressive brands that are making sourcing from regenerative farms and ranches a cornerstone of their business. Doing this cultivates a new data-driven democracy for consumers as well. According to The New Mindful Consumer study by The Innovation Group at J. Walter Thompson Intelligence, 89% of consumers say they care personally about protecting the planet and 70% would be willing to pay more for products that protect the environment. Now, for the first time in history, they can pick up a package and know with empirical data to back it up that a product was grown in a way that improves the environment. This brings new dimensions of possibility to the old cliche of vote with your dollar.

Business can be a force for good, but to deliver on that it needs to prioritize outcomes and play the role of sending the right signals through the marketplace. In an era where brands are being asked to do more impact accounting, set increasingly aggressive CSR targets, align with UN Sustainable Development Goals, etc., established leader brands from the natural and organic space are seeing the opportunity to gain first-mover advantages. Smaller and younger brands are also seeing the opportunity to garner market share by aligning with consumers on a broader array of their values. The time is now, the exam is tomorrow morning and we havent done too well on previous tests. Lets do this!

Chris Kerston co-leadsLand to Market, a program of the Savory Institute thatconnects agricultural producers with progressive brands.

Read more:

Why regenerative agriculture must be measured - New Hope Network

It takes something special to be even more astounding than a Matt Gaetz alibi, but Judy Batalions new book, The Light of Days, achieves that and much, much more.

Even the books subtitle The Untold Story of Women Resistance Fighters in Hitlers Ghettos doesnt do justice to the amazing tales recounted in this labor of love from the Canadian-born New Yorker.

The 20 young Jewish women she spotlights lived remarkable lives during World War II, and its easy to see why Steven Spielbergs Amblin Entertainment snapped up the film rights at manuscript stage in 2018.

Batalion, 44 this month, is currently co-writing the screenplay, and while no director is currently attached, many of the true stories here feel like something from the mind of Quentin Tarantino (think Inglorious Basterds) rather than a more traditional Holocaust drama like Schindlers List.

Take, for example, the story of Bela Hazan, a fearless 19-year-old from southeastern Poland who took a job working in, of all places, a Gestapo office. This was the perfect cover for her to act as a courier for a rebel group from the Dror youth movement, smuggling news bulletins, money and weaponry across Nazi-occupied Poland. (Dror and other youth movements like Hashomer Hatzair became a de facto Jewish resistance network in the war.)

Then theres Renia Kukielka, who was just 14 at the start of the war but went on to become a crucial courier ferrying messages between ghettos. Or Zivia Lubetkin, who was in her mid-20s when she played a key yet long overlooked role in the Warsaw Ghetto Uprising of April 1943 as part of the Jewish Fighting Organization (also known by its Polish acronym, the ZOB).

The Light of Days the books title comes from a line written by a young Jewish girl for a ghetto song contest is both a profoundly moving and breathtaking read, full of tragic and audacious stories. Yet it also provokes anger that it has taken some 75 years for these stories to themselves see the light of day and for these acts of heroism finally to be acknowledged.

Some of the young women Batalion showcases were partisans, literally fighting the Nazis deep within the forests of Eastern Europe. Many others operated as couriers, bringing news of Nazi atrocities to Polands 400-plus ghettos or smuggling in munitions, cash and even fighting spirit.

Why were women chosen for these tasks? Obviously, there was no way for the Nazis to physically prove a woman was Jewish. But equally importantly, many were more familiar with Polish culture than their male peers and could blend in more easily. These were educated young women who could think on their feet and pass as their Aryan compatriots.

Ode to guns

The Light of Days begins with the wars most celebrated Jewish resistance fighter, Hannah Szenes. It was while researching a story on her, at the British Library in London in the spring of 2007, that Batalion discovered a very dusty blue volume among the small pile of books about the volunteer parachutist.

She almost set it aside, but the historian in her forced her to pick it up and examine it. It was an unusual book for the British Library to hold, since it was in Yiddish. But that wasnt the only unusual thing: Batalion actually speaks Yiddish too, so was able to read the 1946 book, called Freuen in di Ghettos (Women in the Ghettos).

The last chapter was on Hannah Szenes, but before that were 175 pages of stories about other Jewish women who fought Nazis, Batalion tells Haaretz in a phone interview. The chapters had titles like Ammunition and Partisan Battles, and in one part there was an ode to guns, she recalls. It was so not what I expected, and so foreign to the Holocaust narrative I had grown up with. It really startled me.

Batalion comes from a family of Polish-born Holocaust survivors and grew up in a tight-knit Jewish community in Montreal, but says much of her early life was an attempt to run away from that. Hence, she found herself in London, performing stand-up comedy and working in the art world, but with questions gnawing away about her Jewish heritage.

But 2007 wasnt the right time for her to emotionally commit to such a mentally exhausting project. The last place I wanted to be at that time in my life was spending my afternoons in 1943 in Warsaw emotionally, socially, intellectually, she recalls. To write this kind of book, I would have to sit with dozens, even hundreds, of these testimonies, and I wasnt ready to do that until later in my life.

Still, Batalion applied for and received a grant to translate Freuen into English, which took about five years (It was a very complicated translation because, first of all, my Yiddish was rusty I dont use Yiddish that much in my daily life. It was also a more Germanic Yiddish, and I grew up with a more Polish Yiddish). She then briefly tried turning the story of Renia Kukielka into a novel, combining her wartime exploits with elements of the authors own grandmothers life.

And finally, in 2017, it was my literary agent who asked me, Wait, what? This really happened? She was the one who told me, You have to write this as a nonfiction book. Its very important to tell the true story, Batalion recounts. And thats how we find ourselves in the rare position of having to praise an agent for their efforts on our behalf.

Freuen was just the starting point for The Light of Days, though. That source material was like a scrapbook, Batalion says, comprising clippings from different newspapers, obituaries, speeches and memoirs about female fighters from Jewish youth movements. Her own extensive research included revisiting numerous wartime sites across Poland, reading and watching whatever testimonies existed, and interviewing the families of the women who survived the war.

But the biggest initial challenge was to work out the chronology of events and how lots of separate stories might mesh together. It took me about six months to do a rough first draft, she says. Im writing history out of memoir, so I had to put together what happened, and when. I was working with personal stories: You can have a whole memoir that takes place in one week and the rest of the war takes up one page, so I had to figure out how these stories worked together.

The people who had survived, or had survived long enough to write about their experiences, were characters that I could focus on, because they had left more detailed, robust stories, she explains.

Then there was the small matter of trying to verify stories that havent been told in nearly 80 years, if at all, and were sometimes written when typewriters, pens and paper werent exactly easy to access.

The book has 65 pages of endnotes and a lot of them say, I took this from this section and this from this, and this memoir said this and in this testimony it said something a little different, Batalion says. I tried to piece together stories, and a lot of times the details did conflict what happened in one account isnt exactly the same as in another account. But the accounts often refer to the same events, which was also exciting as a researcher. Theyre all talking aboutthat day in 1942. I had to decide what version seemed the most historically accurate and made sense.

Another challenge in a book like this is getting the right balance between the heroes and martyrs, to use the Hebrew term for Israels Holocaust Remembrance Day which, significantly, occurs on the anniversary of the start of the Warsaw Ghetto Uprising.

There were a lot of balances to get right, Batalion notes. Im writing here in the U.S., where a huge percentage of the millennial population doesnt even know what Auschwitz is, she says, referring to the 2018 survey that found two-thirds of millennials had never heard of the death camp. Its very tricky to tell a story about the Holocaust, because I want to explain the deeply horrific nature of this genocide, but I also want to tell a story of the people that fought it.

A scholar who wrote a book about humor in the Holocaust wrote, If you want to write about humor in the Holocaust, the danger is that it seems like the Holocaust wasnt that bad. This resonated with me. I didnt want to make it sound like there was a massive Jewish army who was fighting the Nazis. This was a horrific genocide, and these were teenagers who tried to organize to overcome.

The author didnt make life easy for herself by choosing to relate the stories of tens of different women (the film, by necessity, will have to focus on a couple of leading characters), and Batalion says this was her most difficult writing decision. This wasnt a story of just two or three women this was a movement of organized resistance across the country that involved hundreds, if not thousands, and it was important that that came across, she explains.

Strong sense of instinct

The authors research uncovered more incredible resistance stories than she ever could have imagined, but I wonder if she found any common traits among these young women to help explain their apparent fearlessness.

You know, Ive thought about this a lot, she says. I think their bold and savvy behavior was shaped by their training, by their youth movements and how they were educated but I also think many of these women had a very strong sense of instinct, and followed it. Im always obsessed with people that I feel have what I lack.

She recounts a meeting with Renia Kukielkas family in Israel a few years ago. They said to me, just in passing, Renia wasnt someone who, when she crossed the street, would look left and right, left and right. And that stayed with me, because I am someone who looks left and right, left and right, left and right. I think many of these rebels had strong impulses and trusted their gut and just moved.

The Light of Days conjures up many indelible images: women hiding razor blades in their hair; secret libraries and makeshift weapons labs being established in ghettos; female couriers donning layers of skirts to hide contraband in the folds; and young women determined not to go like sleep to the slaughter, to quote Jewish partisan leader Abba Kovners resistance mantra.

Two other things leap out at you. One is to be reminded of the sheer scale of the Nazi killing machine, with the Germans establishing over 400 ghettos across Poland alone. For Batalion, its both the big numbers and the smallness of the places that overwhelm. Over 400,000 Jews were forced to live in the Warsaw ghetto alone. Thats a huge number. I was also shocked by the scope of resistance participation: Over 90 European ghettos had armed Jewish underground movements. Id had no idea.

And then, on the other hand, theres the smallness. When you go to these towns and walk through the streets of former ghettos, theyre just small-town streets. Even some of the camps that I visited, theyre very human in size in my head they loomed so large. The Gestapo headquarters [in Warsaw] is a four-story building, its so regular which is equally troubling, in a way.

The second thing that strikes you is the joie de vivre exhibited by so many of these young Jews, despite or perhaps because of the horrors of everyday ghetto life. Indeed, a recurring question as you read the book is, when did these people ever sleep?

Batalion: Every testimony I read, every memoir I read, was just so full of action they were so alive. These were stories of constant activity, and they drew me in. These women were literally jumping off trains, running between towns, getting dressed up, dyeing their hair. These were stories with so much action, and I think that also just changed the tone of the Holocaust narrative for me. Its so different from the more staid narrative I had been exposed to.

It is also impossible not to read The Light of Days and see it as the current Polish governments worst nightmare in light of its controversial, some would say revisionist, stance regarding the role its citizens played in World War II: a book that presents the Holocaust in all its complexities, depicting some non-Jewish Poles as heroes but many others as aiding and abetting the Nazis or committing their own atrocities.

The good news is that The Light of Days will be published in Poland next year, so locals will be able to make up their own minds, while Batalion has only good things to say about the Poles who assisted her in the writing process.

My only reactions have been from people who helped me do research in Poland translators, research assistants, drivers, fixers and I honestly felt that they were as interested in this story as I was, she says. They were so passionate about it, this was so important to them. To them, this is Polish history; this is their story too. For me too, this is a Polish history book.

I made fascinating connections in Poland, mainly with young people in their 20s and 30s. At my Polish publisher, I was saying casually that all four of my grandparents were from Poland and they laughed, saying, Youre more Polish than any of us! I have a fraught and complicated relationship to Poland, but I was taken by how passionate these young Poles were about my project.

Three lines in history

Polish historian Emanuel Ringelblum, the noted chronicler of Warsaw ghetto life, is quoted in Batalions book describing how the women put themselves in mortal danger every day to carry out the most dangerous missions. Nothing stands in their way. Nothing deters them. Yet his prediction that the story of the Jewish women will be a glorious page in the history of Jewry during the present war turned out to be far from accurate.

Why has it taken so long for these stories to finally be told and for these women to get their three lines in history, as one young ghetto activist puts it? Batalion has her own theories.

The story of why I dont know this story is to me as interesting as the story itself, she says. There are many reasons why this tale disappeared some of them have to do with the Zeitgeist and the interests of the times; some of them have to do with politics. And some of them are very personal. These women didnt tell their story. Or they told them right after the war, like Renia, and that was it. The telling was in a sense the therapy, or part of the therapy, and then they had to move on. It was so important to start afresh. As I mentioned in the book, some of these women werent believed. Some of them were accused of leaving their families or sleeping their way to safety. Many of these women suffered terrible survivors guilt.

So, things were silenced for many reasons, and a lot of it had to do with these women feeling very determined to create families, to create a new generation of Jews and they didnt want to hurt them. They wanted their children to be healthy and happy and normal.

As her own toddler starts screaming in the background, demanding her attention, Batalion just has time to express her hopes for a book 14 years, or perhaps several lifetimes, in the making: I just want people to know these stories. I want people to know their legacy. I want people to know the names of these women who fought against all odds for our collective justice and liberty.

The Light of Days: The Untold Story of Women Resistance Fighters in Hitlers Ghettos is out on Tuesday, published by William Morrow, priced $28.99. The book will be published in Hebrew by Yediot. Visit judybatalion.com/events for details of online talks connected to the book.

See the article here:

The young Jewish women who fought the Nazis and why youve never heard of them - Haaretz

Conversations around AI and ethics may have started as a preoccupation of activists and academics, but now prompted by the increasing frequency of headlines of biased algorithms, black box models, and privacy violations boards, C-suites, and data and AI leaders have realized its an issue for which they need a strategic approach.

A solution is hiding in plain sight. Other industries have already found ways to deal with complex ethical quandaries quickly, effectively, and in a way that can be easily replicated. Instead of trying to reinvent this process, companies need to adopt and customize one of health cares greatest inventions: the Institutional Review Board, or IRB.

Most discussions of AI ethics follow the same flawed formula,consisting of three moves, each of which is problematic from the perspective of an organization that wants to mitigate the ethical risks associated with AI.

Heres how these conversations tend to go.

First, companies move to identify AI ethics with fairness in AI, or sometimes more generally, fairness, equity, and inclusion. This certainly resonates with the zeitgeist the rise of BLM, the anti-racist movement, and corporate support for diversity and inclusion measures.

Second, they move from the language of fairness to the language of bias: biased algorithms, as popular media puts it, or biased models as engineers (more accurately) call it. The examples of (allegedly) biased models are well-know, including those from Amazon, Optum Health, and Goldman Sachs.

Finally, they look for ways to address the problem as theyve defined it. They discuss technical tools (whether open source, or sold by Big Tech or a startup) for bias identification, which standardly compare a models outputs against dozens of quantitative metrics or definitions of fairness found in the burgeoning academic research area of machine learning (ML) ethics. They may also consider engaging stakeholders, especially those that comprise historically marginalized populations.

While some recent AI ethics discussions go beyond this, many of the most prominent dont. And the most common set of actions that practitioners actually undertake flows from these three moves: Most companies adopt a risk-mitigation strategy that utilizes one of the aforementioned technical tools, if theyre doing anything at all.

All of this should keep the stewards of brand reputation up at night, because this process has barely scratched the surface of the ethical risks that AI introduces. To understand why this is, lets take each of these moves in turn.

The first move sends you off in the wrong direction, because it immediately narrows the scope. Defining AI ethics as fairness in AI is problematic for the simple reason that fairness issues are just a subset of ethical issues youve just decided to ignore giant swaths of ethical risk. Most obviously there are issues relating to privacy violations (given that most of current AI is ML, which is often powered by peoples data), and unexplainable outputs/black-box algorithms. But there are more. For example, the primary ethical risk related to AI-powered self-driving cars isnt bias or privacy, but killing and maiming. The ethical risk of facial recognition technology doesnt end when bias has been weeded out from the model (of which there are a number of examples); a non-biased facial recognition software still enables surveillance by corporations and (fascist) governments. AI also requires vast amounts of energy to power the computers that are training the algorithms, which entails a shocking degree of damage to the environment. The list of ways a company can meet with ethical disaster is endless, and so reducing AI ethics to issues involving fairness is a recipe for disaster.

The second move further reduces your remit: Issues of bias are a subset of issues of fairness. More specifically, issues of bias in the context of AI are issues of how different subpopulations are treated relative to others whether goods, services, and opportunities are distributed fairly and justly. Are the job ads placed in such a way that they are just as likely to be seen by the African-American population as by the white population? Are women applying for a job just as likely to have their resumes lead to an interview as a mans resume? The problem with this approach is that issues of fairness extend beyond issues of fair distributions of goods to various subpopulations.

Most obviously, there are issues about what any individual deserves independently of how others are treated. If Im torturing you, and you protest, it would hardly justify my actions to say, Dont worry, Im torturing other subpopulations at equal rates to the population of which youre a member. The entire category of human rights is about what every person deserves, independently of how others are being treated. Fairness crucially involves issues of individual desert, and a discussion, let alone a risk-mitigation strategy, that leaves this out is the more perilous for it.

The final trouble for organizations arrives in the third move identifying and adopting bias-mitigation strategies and technical tools. Organizations often lean on technical tools, in particular, as their go-to (or only) meaningful instrument to ferret out bias, as measured by quantitative definitions of fairness found in recent computer science literature. Here, we run into a raft of failures at ethical risk mitigation.

First, those two-dozen-plus quantitative metrics for fairness are not compatible with each other. You simply cannot be fair according to all of them at the same time. That means that an ethical judgment needs to be made: Which, if any, of these quantitative metrics of fairness are the ethical/appropriate ones to use? Instead of bringing in lawyers, political theorists, or ethicists all of whom have training in these kinds of complex ethical questions these decisions are left to data scientists and engineers. But if the experts arent in the room, you cannot expect your due diligence to have been responsibly discharged.

Second, these tools standardly only kick in well into the development lifecycle. Because they measure the output of AI models, theyre used after data sets have been chosen and models have been trained and a good deal of resources have been devoted to the product. It is then inefficient, not to mention unpopular, to go back to the drawing board if a bias problem is detected that cannot be solved in a fairly straightforward way.

Third, while the search for a technical, quantitative solution to AI ethics is understandable, the truth is that many ethical issues are not reducible to quantitative metrics or KPIs. Surveillance is a problem because it destroys trust, causes anxiety, alters peoples behavior, and ultimately erodes autonomy. Questions about whether people are being treated respectfully, whether a product design is manipulative or merely giving reasonable incentives, whether a decision places a burden on people that is too great to reasonably expect of them, these all require qualitative assessments.

Fourth, these technical tools do not cover all types of bias. They do not, for instance, ferret out whether your search engine has labeled Black people gorillas. These are cases of bias for which no technical tool exists.

Fifth, the way these tools measure for bias are often not compatible with existing anti-discrimination law. For example, anti-discrimination law forbids companies from using variables like race and gender in their decision-making process. But what if that needs to be done in order to test their models for bias and thus influences the changes they make to the model in an effort to mitigate bias? That looks to be not only ethically permissible, but plausibly ethically required as well.

Finally, as regards engaging stakeholders, that is generally a good thing to do. However, aside from the logistical issues to which it gives rise, it does not by itself mitigate any ethical risks; it leaves them right in place, unless one knows how to think through stakeholder feedback. For instance, suppose your stakeholders are racist. Suppose the norms that are local to where you will deploy your AI encourage gender discrimination. Suppose your stakeholders disagree with each other because, in part, they have conflicting interests; stakeholders are not a monolithic group with a single perspective, after all. Stakeholder input is valuable, but you cannot programmatically derive an ethical decision from stakeholder input.

My point here isnt that technical tools and stakeholder outreach should not be avoided; they are indeed quite useful. But we need more comprehensive ways to deal with ethical risk. Ideally, this will involve building a comprehensive AI ethical risk-mitigation program that is implemented throughout an organization admittedly a heavy lift. If a company is looking for something to do in relatively short order that can have a big impact (and will later dovetail well with the bigger risk-mitigation program), they should take their cues on ethical risk mitigation from health care and create an IRB.

In the United States, IRBs in medicine were introduced to mitigate the ethical risks that arose and were commonly realized in research on human subjects. Some of that unethical conduct was particularly horrific, including the Tuskegee experiments, in which doctors refrained from treating Black men with syphilis, despite penicillin being available, so they could study the diseases unmitigated progression. More generally, the goals of an IRB include upholding the core ethical principles of respect for persons, beneficence, and justice. IRBs carry out their function by approving, denying, and suggesting changes to proposed research projects.

Comparing the kinds of ethical risks present in medicine to the kinds present in AI is useful for a number of reasons. First, in both instances there is the potential for harming individuals and groups of people (e.g. members of a particular race or gender). Second, there exists a vast array of ethical risks that can be realized in both fields, ranging from physical harm and mental distress to discriminating against protected classes, invading peoples privacy, and undermining peoples autonomy. Third, many of the ethical risks in both instances arise from the particular applications of the technology at hand.

Applied to AI, the IRB can have the capacity to systematically and exhaustively identify ethical risks across the board. Just as in medical research, an AI IRB can not only play the role of approving and rejecting various proposals, but should also make ethical risk-mitigation recommendations to researchers and product developers. Moreover, a well-constituted IRB more on this in a moment can perform the functions that the current approach cannot.

When it comes to building and maintaining an IRB, three issues loom large: membership of the board, jurisdiction, and articulating the values it will strive to achieve (or at least the nightmares it strives to avoid).

To systematically and exhaustively identify and mitigate AI ethical risks, an AI IRB requires a diverse team of experts. You will want to place an engineer that understands the technical underpinnings of the research and/or product so the committee can understand what is being done and what can be done from a technical perspective. Similarly, someone deeply familiar with product design is important. They speak the language of the product developers, understand customer journeys, and can help shape ethical risk-mitigation strategies in a way that doesnt undermine the essential functions of the products under consideration.

Youll also want to include ethics-adjacent members, like attorneys and privacy officers. Their knowledge of current and potential regulations, anti-discrimination law, and privacy practices are important places to look when vetting for ethical risks.

Insofar as the AI IRB has as its function the identification and mitigation of ethical risks, it would be wise of you to include an ethicist, e.g. someone with a Ph.D. in philosophy who specializes in ethics, or, say, someone with a masters degree in medical ethics. The ethicist isnt there to act as kind of priest with superior ethical views. Theyre there because they have training, knowledge, and experience related to understanding and spotting a vast array of ethical risks, familiarity with important concepts and distinctions that aid in clear-eyed ethical deliberation, and the skill of helping groups of people objectively assess ethical issues. Importantly, this kind of risk assessment is distinct from the risks one finds in the model risk assessments created by data scientists and engineers, which tend to focus on issues relating to accuracy and data quality.

You may also find it useful to include various subject-matter experts depending on the research or product at hand. If the product is to be deployed in universities, someone deeply familiar with their operations, goals, and constituencies should be included. If it is a product to be deployed in Japan, including an expert in Japanese culture may be important.

Lastly, as part of an effort to maintain independence and the absence of conflict of interests (e.g. members looking for approval from their bosses), having at least one member unaffiliated with your organization is important (and is, incidentally, required for medical IRBs). At the same time, all members should have a sense of the business goals and necessities.

When should an AI IRB be consulted, how much power should it have, and where should it be situated in product development? In the medical community, IRBs are consulted prior to the start of research. The reason there is obvious: The IRB is consulted when testing on human subjects will be performed, and one needs approval before that testing begins. When it comes to authority, medical IRBs are the ultimate authority. They can approve and reject proposals, as well as suggest changes to the proposal, and their decisions are final. Once an IRB has denied a proposal, another IRB cannot approve it, and the decision cant be appealed.

The same rule should apply for an AI IRB.

Even though the harm typically occurs during deployment of the AI, not research and product development, theres a strong case for having an AI IRB before research and/or product development begins. The primary reason for this is that its much easier and therefore more inexpensive and efficient to change projects and products that do not yet exist. If, for instance, you only realize a significant ethical risk from a potential or probable unintended consequence of how the product was designed, you will either have to go to market with a product you know to be ethically risky or you will have to go through the costly process of reengineering the product.

While medical IRBs are granted their authority by the law, there is at least one strong reason you should consider voluntarily granting that degree of power to an AI IRB: It is a tool by which great trust can be built with employees, clients, and consumers. That is particularly true if your organization is transparent about the operations even if not the exact decisions of the IRB. If being an ethically sound company is at the top of the pyramid of your companys values, then granting an AI IRB the independence and power to veto proposals without the possibility of an appeal (to a member of your executive team, for instance) is a good idea.

Of course, that is often (sadly) not the case. Most companies will see the AI IRB as a tool of risk mitigation, not elimination, and one should admit at least the possibility if not the probability of cases in which a company can pursue a project that is ethically risky while also highly profitable. For companies with that kind of ethical risk appetite, either an appeals process will have to be created or, if they are only minorly concerned with ethical risk mitigation, they may make the pronouncements of the board advisory instead of required. At that point, though, they should not expect the board to be particularly effective at systematically mitigating ethical risks.

Youve assembled your AI IRB and defined its jurisdiction. Now youll need to articulate the values by which it should be guided. The standard way of doing this is to articulate a set of principles and then seek to apply those principles to the case at hand. This is notoriously difficult, given the vast array of ways in which principles can be interpreted and applied; just think about the various and incompatible ways sincere politicians interpret and apply the principle of fairness.

In medical ethics and in the law, for that matter decision-making is usually not guided by principles alone. Instead, they rely on case studies and precedent, comparing any given case under investigation to previous cases that are similar. This allows your IRB to leverage the insights brought to bear on the previous case to the present case. It also increases the probability of consistency in the application of principles across cases.

Progress can be made here by articulating previous decisions senior leadership made on ethical grounds prior to the existence of the IRB. Suppose, for instance, the IRB knows senior leaders rejected a contract with a certain government due to particular ethical concerns about how the government operates generally or how they anticipated that government would use their product. The reasoning that led to the decision can reveal how future cases ought to be decided. In the event that no such cases exist and/or no such cases have been disclosed, it can be useful to entertain fictional examples, preferably ones that are not unlikely to become real examples in the future, and for the IRB to deliberate and decide on those cases. Doing that will ensure readiness for the real case when it arrives on their doorstep. It also encourages the cool objectivity with which fictional cases can be considered when no money is on the line, for instance to transfer to the real cases to which theyll be compared.

***

We all know that adopting an AI strategy is becoming a necessity to stay competitive. In a remarkable bit of good news, board members and data leaders see AI ethical risk mitigation as an essential component of that strategy. But current approaches are grossly inadequate and many leaders are uncertain how to develop this part of their strategy.

In the absence of the ideal a widespread commitment to creating a robust AI ethical-risk program from day one building, maintaining, and empowering an AI IRB can serve as a strong foundation for achieving that ideal. It can be created in relatively short order, it can be piloted fairly easily, it can be built on and expanded to cover all product teams and even all departments, and it creates and communicates a culture of ethics. Thats a powerful punch not only for AI ethics, but for the ethics of the organization more generally.

Follow this link:

If Your Company Uses AI, It Needs an Internal Review Board - Harvard Business Review