I have the video open on my computer and it started playing my own words back to me. It was vy confusing to stop our next spker is the author of black software e internet and Racial Justice. Is the Vice President of faculty engagement and developmenat new york university. And professor media, culture and communication at nyu Steinhardt School of the father for the center of critical race d indigenous studies based on finer stomach final speaker will be frank aauthor of the new laws of robotics defending Human Expertise in the age of aib stopped his professor of law atbrooklyn law school at affiliate llow at the Yale Information Society project and a member of the american law institute. Thank you. Im so happy to be on the panel with two authors i reay appreciate and with that topic im sure we all have so much to say about. There were many periods in time that i could look at as representative of what my ideals or best opportunities, i wanted to focus attention to 10 years ago, 12 years ag the time of the great recession, the last jor financial crisis and how it was a unique opportunity for Tech Companies to thrive. There were a number of moments familiar to a lot of people, and i imagine that is about the time the smart phone took off, that was about the time that because there waa financial crisis and because i had implications with not just wall street but real estate, there was a drive to invest elsewhere in certain shiny new industries popping up and location that is not in new rk. So with all of the elements coming together, another element was a lot of the lack of scrutiny thathese companies got, perhapsot on the level of the user because plenty of users abuse technologies has issues, had issues with facebook, but the opportunity to find a platform to discuss their problems, to discuss the many consequences they were feeling as users, there wasnt mh press about, there wasnt a lot of critical press. It was very difficult to place critical stories about technology in newspapers and magazines. And therwas just these think tankand academic researc ceers and its wonderful that they are all taking off but that theyve taken off in the past 10 years, this is relatively new. I want people to take a look back at the founding of these companies to see what problems perhaps where they are from the beginning so that hearing back considerably so that some ideas i wanted to throw out to begin. Great, charlton. Excellent, i think that is a great segue to some of my remarks and maybe will take a little bit more time and if the audience wi ai was thinking last nig about i really wanted to say and bring forward in this discussn to lead with i had th moment where i had the feeling i had written thes things before ai was collecting my thoughts and then i looked back into the book and looked at my intruction and found, i think i said it actually here and said it better. I want to read a short rt of the introduction to black software contexand then within two or three provocations that really lead us back to exactly where joannes comments ended, which is about, what do we do now and how do we proceed. The book of cose is black software, for me black softwa, hundreds of myriad ways we mobilize Computing Technology refers to t programs we desire and designed computers to run, refers to who designed the program for what purposes d what or whom becomes this object or data. Fers to how and how well the computer performs a task for which it was programmed. In this book i tell two black Software Stories the book i organized the book 1 and book 2. Is is where a group called the vanguard that i ite about begins to emerge. These are black hobbyists and entrepreneurs, digital organizersevangelists, activists, knowledge brokers and they positioned black folk black content, black culturto occupy the leading edge of the internet popular social deponent. So they e collective stories begin around the mid70s continue into the 80s and 90s during the advent of personal computing and early days of computer networking and air stories extend to the World Wide Web first and. Com booms first bus, the story demonstrate how black people have taken talkhnology abusing the technology to further our own purposes, our persal communal and political interests. But blk software is also a story about how Computing Technology was built and developed to keep america docile and in its place disproportionately disadvantaged,ocked up and marked for death. This is a story i unravel in book 2 and speaks tothem in various ways of those in power use computing of technology to destroy and nullify black agency and nullified black peoples hopes and dreams, aspirations and Human Potential and political interests. Limiting the heights we are meant to achieve. Book 2 begs in 1960 and really has to do with a group people i refer to as the commitemen and president Lyndon JohnsonsCrime Commission in 1965 who built new computing systems that were really directed at what was then seen as america preeminent problem, that was the problem of race and the problem of blackness. I and the introduction with these words in between the two versions of black software, the kind that positively impacts black peles lives and the kinds that stroy them, lives of both significant question which is not about recently popularized concepts like computer bias or fa algorithms or platform and equality or ethics. The queson goes to the heart of the matter that the concepts skirt around. Will our current and future technological tools enable us to outrun White Supremacy . Ter all, this is not just our countrys founding principle, its also the core programming that preceded and animated th birth, development, and uses of our computational systems. Im often asked today, can we use technology to significantly advance the cause of rial justice and the aner is, only if we come to grips with the fact that our current technological environment and infrastructure was designed to support these efforts of Racial Justice and civil rights. I think that whathat means is that its incumbent ons to think about what it means, what it would mean to dismantle and rebuild our technical logical environment and infrastructure from the ground up that is inherently or at least more inherentlybent toward the cause of racial juice. So what does that mean . What does that look like thats a question that is very much a part of our future. Thank you. Thanks so much frank. Hank you christina, thank you so much to joan and charlton i think these projects are at the center of how we can rethink technology for soci justice. I think, in my book new laws of robotics, i will get to the new laws and a second but i want to give a reallconcrete example of w ai robotics might affect our everyday lives. We could use an old form of Job Interview to walk in and have an interew with someone to talk abouthings we fill out a form etc. And then either hired or not hired. There are now firms developing ai fial analysis and facial Recognition Systems that will parse microsecond by crosecond every emotion you ow, i shouldnt say emotion, i should say every w in which her face moves. How fast you blink. Do you smile or not . Things like that. Im sure weve all had selfconsciousne in dealing with the zoom universe in exactly this way. Whats really troubling about it is that these firms are saying, we can look at your existing employees, your top performers, and then see who and society matches the way that they speak, the way they sound, things like that, i think theres a uple of responses to that that i think, there is an initial response to be worried about it and part of my book is to arculate why we should bworried and the how to translate those worries into policies to improve things. I think of a first wave of worry is exactly along the lines of charless points about these systems being systematicly disadvantaging to minority ties groups in society. I think have to really watch carefully for this impact here. If we allow firms to use ai to essentially replicate themselves or replicate along certain lines, is it about really finding high peormance or about simply replicating the existing traits of the cuent . There is a second way that goes into the ideas of joae thinking what how you consider the computer to say that we might want to question the use of thatechnology altogether because we might find it safe to humanize it. Want to be evaluated on the basis of some personal interaction, i deserve that, thats the minimum i deserve. We have 10,000 applications for 20 slots, how are we supposed to screen them . They already use things like resume screeners, resume automated natural language processing, that can be troubling but i think its less troubling for the sort of thing that is analyzing peoples faces in ways that people really have no control over. Ople dont have any control over whether they blink 10 microseconds longer than they ould in order to get in position. And thinking about ai robotics, thats an ai application, later in the book i also try to develop a more hopeful vision for robotics as complements to professionals rather an substitutes. There is lots of places where rots can substitute human nature well, and eventually will do well. In many professions you really want to have the complementarity. Thats the first law of robotics, have ai robotics consummate professionals rather than replace them. The second is to avoid having the stems trying to counterfeit humanity so if theres box onli, dont allow them to pretend to be equal is not this is also rai serious civil rights concerns because you are right when groups that have put fake minority online that say we love president trump,thats a deeply troubling thing to see those sorts of bs impersonating people. My third thing about robotics is to stop arms ces and robotics. Theres many other forms of arms races think ai is comforti to. My fourth law of robotics is to require attribution for any ai robotics system. In the book but i essentially do is in the fields of law, medicine, journalism, education, and a few other fields, try to develop those and say whatould our laws and policies look like if we held riously to these principles in the new laws orobotics. I look forward to the discussion today. Thank you. Thanks everyone, if anyones watching this contemporaneously, i have my ey on the chat if there any questions. My questiois more aimed at the books of charlton and frfrank, part of the reason may be the reason that you see biased in technology is becaus its made by humans who are biased. So that raises the question of, does it help to focus more to say you shouldnt focus so much on technology we should focus on the people or people should ve more decisionmaking power because they are all from the same cause of the technology side. Viously the systems work differently i wawondering if both of you could speak to how you envision writinghe ship when the underlyinproblem zones arent necessaril problems are to abpeople are technology but the attitudes people that create both. That is a great question and one that i think about that do agree that when we think about where and what the fix is that its more with human beings than it is of technology and i think one of the thingthat we started to get geourselves into is thking nacvely that if we can abi think we make a miake thinking about things that way and i think one of the fundamental problems people are designing these kinds of systems, thinking about with franks example the facial requisition in the intervw and employment context how many of those folks are very well versed in the history and context of the face and what that means psychologicalland what historically those connections tween facial attributes anracial biases stereotypes etc. And then what kind of probms might that cause in aechnology that uses at to then infer something about your qualificions. I think that is where the disconnect is both in terms of dealing with individuals attitudes but also Historical Context around race, racism, all the things thateople have no idea about is easy to replicate when all you are thinking about is, i want to make a re efficient system to help me t through these resumes. I reay like that point about the history, i ink its something that is ideally going to be part of Computer Science education of the fure is to have a much larger role for understanding the hiory of the social context. I think when we look at the psychology of things li facial analysis program, its often really partial to a behaviorist model of an bottling humans as opposed to things thamight be more phenomenological or open to peoples lived experience in thnarrative about their exrience. Think to get to your question, christin part of my hope in writing on thitopic is to try to make it so people dont either we pick a robot or we have humans doing it but weve got this sociotechnical configuration of humans in robot anhow do we make that so it serves this process were ideally we n look at the way humans do statistically or things like bias, report back to people, you appear to be acng a certain way that is didvantaging certain members of community d then reacting but i want to see humans at the center of the process because i worry a bit about that happening its entirely Automated Press given the record of big Tech Companies that do the automation. Joanne, your comments he mentioned that if you go back a little bit in time it was hard to place essays orarticles that were critical on emerging technology, big Tech Companies, now that narrative is very much changing and changed. You would think the biggest villains in our society are often so of these very large companies. What do you attribute to th change and can learn something from it . Its funny because today is actually the 1h anniversary of the release of thissocial network im sure eveones heard of it the hollywood film about the foundingf facebook. I dont think its necessarily the best critism of facebook but it was hollywood film depiction at was up for academy award. That was 10 yes ago. Pretty commoy when i see news articles or Opinion Pieces now day, i see a very common beginning that will be like, 10 years ago Everyone Wants facebook. Everyone used to love facebook now they are afraid of it. This is a really common structurand i understand it as a writer because its a rhetorical technique to give your opponent credit before you go into the criticism. But the reason i hope people watching this who them selves mighbe a person to speak on these issues, keep in consideration that that is a narrative that marzuckerberg has crafted about facebook. He himself is someone who when he gets interviewed he says, all of this is very new criticis you all used to love us. So the ways that because a few people were documenting it back th, even though it was aba few people in ntrast with the number of users facebook had, i dont have a number off the top, even 10 years ago its user base, certainly would have counted as a major city in the united states. That something i want people to be aware of because if we belie in this idea of a beautiful past, thats very easy to get back to you, we have very surface level changes we can g back to that good place these Tech Companies were about what i would like others to consideinstead is just why do we have a company like facebook. Google search is very useful. So what cebook offers, they will tell you we are great becauswe let you connect with people. We let you do th, this, this, that is something the internet set self alrea let you do, that dont have to have the commercialized functions and targeted ads that weve placed in the middle. Everything that it offers you can do with the internet already. Also, everything it offers, some of its very useful functions say you need a facebook account to t updates on your School Committee meeting or your church, things related to your church. Wouldnt that all be better and decentralized space clear you are the moderator where the community is managing it to mackey isnt all of this water . I think about the commity newspapers thawere maybe they werent riving 10 years ago. When you think about the way that a commity newspaper might have had lot of alerts about yourchool, about your representatives, people from your tow now that facebook has taken over, youve got people lookingmorelobal when they look glal they get embroiled in things like conspiracies. I dont think you would have q1 unnecessarily in a committee newspaper 10 years ago. Im sure it couldve happened but it would have been been e same way facebk has facilitated. I wanted to wrap itp, i just wanteople to think about what do you actually get fm this and do we need it . Dont think we do. So the rhetorical question you are asking in your comments isnt it better on the community on a communities webpage or Something Like that, at some point you can talk about degree of free will. At some ways a lot of these are chosto represent on facebook instead of more decentralized means. What you attribute those reasons and can learn something from those choices if we think that ultimate those were in the wrong and thinking about some of what i read about in blk software which is this moment, albeit very brief in the mid90s where you had, what i like to call thisanalog to the soci media like facebook and so forth but in the aol wall garden and framework where i think there was still something characteristically and positively different there then the facebook meaning youould have a platform like the net nor thatas a site that was heavily populated by targeted to africanamericans. It gained a following and some commercial appeal and very successful in that way but people knew where they were going to get what they wanted without a filter. You had a Thriving Community like a Community Newspaper i could go therbecause i know, here is where im going to find my people, here is where i connect, get the news th i want and then other sitein the mid90s ecosystem that would help you through those old hyperlinks basically help you travel. If you want this that and the other, i can get you from here to that next site or this next site, they all have things that will be of interest but not having the kind of cative filterthat is a facebook that overwhelmingly sort of shows to you it thinks you should be thinking about at the moment. Im very interested and curio in this sort of a prior moment to look backo as an alternative to Something Like this and the platform youre mentioning. If i could jump in on that to the point about alternative is looking to path not taken, i thought it was really interesting aspect of ur discussion of black voices, that was an aol. I felt like in reading that and think about some of the paths we didnt take, i feel it part of the story is prature automation of media. These companies, facebook, google, they are like put up everything and we will curate evything. To me its like, whawould be better ultimately would be t have more people in the more verse spaces and if there was some way to get that goin i wish we had an alternative and maybe cbh was something that helped us not havethat. E extent that you tell the giant players, youe not liable for almost anything that goes on, then the smaller players are able to, the smler players can get pushed out whereas if you have more direct response ability, i think itsinteresting from a legal perspective to think about, at whatoint does the law make stakes and push that direction. I felt the same way also about joannes discussion about the trans rights transcript that was transitioning and they ar on youtube and then their data was used for facial analysis. Its a story where i think a better data regime would stop Something Like that and misuse of the user or misdirection of our media a lenses. Theres a few things and a major component of the answer of all these problems for lack of a better umblla term educion and computer scientists people who argoing to work for companies that enga in automation need to learn more about pshology, history, injustice ethics, in a y that maybe has been less emphasized in education in the past. Or is that not eugh and harder limits on what the law allows or other kinds of things that limit whats possible and push people into less automated directions or less consolidated direction . To start wi a couple of thoughts, i think education is certainly a key component but im not so sure at it is enough because i dont know that it holds the compulsion to do things that absolutely have to be done or must be done. I think back to the transitional moment between the web of the midand late teen 1990s and what became in 2000 later. And particularly the peripheral asian of black cultural articts and content but without black people and without the ability of black people to profit from that which we see from the late 90s and part of the big platforms ther gobbling up or disbanding some ofhese prior assets but we see them over d over whether they pop up on twter or tik tok and so forth. Its the same black cultural element but no attribution of origination of war that intellectual operty comes from and no ability tourn folks to profit, i think there needs to be some legal driver, some policy drivethat helps to moderate what it is we do. If i could just follow uon that point, i think what charltonust said was spot on in terms of the issueeally is ownership. Education is great but thats not going to actually counteract some of the problems we are facing at the moment. Something i also wanted to circle back to an id of abi think with aols involvement, to think about in the 90s the model was cable tv and magazines. L had this idea that if we have atation for black communities like bet, we can capture all their advertising moy. So thats not necessarily the mostltruistic reason but then people are in the room and facebook did not have that. Facebook h its harvard dropouts it had its own a community. They did not have peoplin the room representative of various communities. They made decisions thatere very harmful to various communities. I think ultimately the conversation does have to come back to ownership. Ideally in my mind its decentralized that the community is itswn community that if you have a church oup or, you knowetter than anyone else how to moderate. And how to implement thatis a little bit confusg. Do i think that some education ound privacy and more conversation abouthese issues will benefit these communities . Keep workers, certainly. But i do think sotimes that conversation takes the place of the more structural change than what we have at the moment. Is a really interesting question in relationship to the ai ethi conversation and the revelatory conversation. Do wor a lot that when things are framed as, lets have a really long discussion, its meant to delay ultimate reactions ani think back to the idea of ownership and intellectual property as well as get so interesting to look at what the australian competition and consumer commissi is doing. Its not as targeted to the concerns that we are raising but it is at least saying to google, lets actually think about who is the ultimate creator of the tngs you are monetizing. I remember seeing the discussion of a lot of aspects of bck culture being monetized ry quickly online and there is no re way to for e ultimate creators abi think that is a real aspect of really important aspect of the future is like, how can we readjust that . When i think about the history of music, part of that involved the licensing entities dending licenses or other forms of licenses and competition. I think there are other ways, at has its own troubling history well in terms of the groupseft behind some of the it schemes. I think theres something about nership and input and governance to be in the conversation. A lot of your suggestions, these very compelling sties about how ai doesnt always rk as initially envisioned or designed. I think whenever people hear them they are really moved and see, i see how humans can make different kinds of mistakes, technology can make different kinds of mistakes than humans we need build these safeards in and that Human Technology working together is a way to avoid the mistakes of those groups. When that story is so compling you think the reason everyone is not necessarily completely enthusiastically on board with this because they havent heard about as many of the times when computers have failed to work as we mightve hoped . Or do you think the is still this ratiol exuberance that makes people really want to believe th it will be great and that the computer will fix it and thamakes people more immune to these kind of stories . Is a great question because i think there has been a lot of thusiasm, i remember presenting my first paper on google in the mid to thousands, one member of the faculty i was presenti at said dont you mess with my google [laughter] it works for me, there was not a lot of people it wasnt working for. Im really enthused by two things. One the antitrust hearings i see in the house. I remember going to a panel of abmatt ellison said to us, u need to make this a visceral story weve all heard about data visualization, what about data for seral days asian. Within year or so algorithms oppression, that book me is when the best amples of really telling compelling stories from her original enunters with google and discriminatory search results to thestory at the end of the book about yelp and how yelp helps accerate the decline of classic neighborhood barber or hearsomeone because they were listening to the people that were saying, we are not being treated fairlyetc. I think theres at idea of this role, makingt visceral is really key. I thinboth of charlton and joannes book really do that. Its great. He one thing that countering your comments, joanne, about the reasons decentralization is helpful is, i think, those of us who might have had to work with people with different amounts of text savviness and interest, people want things to be easy, they want tngs to just kind of work. May be for Artificial Intelligence in the media, the search results, its hard to curate information when y have to do the work yourself. Its nice if someone do it for you. Given all of the harms that me from that but a lot of benefits from decentralization but its harder when you are not relying on another major entity or any number of whether its a large corration or ai to help you or anythi along those lines. Hodo we square that circle . I think its a really great question. I can dict the viewers to a great research, a great resource called run your own social. He talks about his experiences ing decentralized platforms. Its not easy, its time consuming, it requires skil, it requires money, it requires all these things that are not as easy as just opening up a laptop. But where we are at now in 2020, we have a range of opportunities so if you have a mmunity right now, if you have say a book club or something li that and you have 50 people y are trying to think, what is the best way to keep in touch . Its nice to have nt time you have to turn your meeting over remotely anyys, when you have a chance to have a community have this conversatn, discuss the pluses andminuses of various ways of solving up in my book i took about one commity, the girl gamers group that begins, they have sub read it, read abre it a aits open to so many people. If you are the only young woman in nebraska a very small town nebraska if you love video games, you dont have look very far to find a community and enter that community. That just the front door. Once you have a posting history you can contacthe moderators so you can have some more personal conversations d say things maybe you wouldnt want on a semi publicpace. These are techniques that a lot of people are working out for their own uses. As we are at this point quite reasonly cynical out the opportities of using a place like facebook i might be easy comes quite a cost. I completely recommend seeing what is reistic enough to maintain but also my offer benefits and privacy or security that these bigger problems can offer. Charlton, had a question for you. Your book was written before the major Racial Justice movements this summer happen. So writi a couple more chapters or tie your book to the recent months, what would you have to say or what would you have to communicate . Thats a great question. A little ambivalent i think theres a couple different strands but most immediately it raises the question about the power and limits of Technology Technology interestingly enough i think very much in the vein of what joanne was speaking of, whose platform are you using . Whose technology a back the questions of ownehip. Can i really mobilize my group on twitter for social action when i know, as we learned in the prior iterations of black like matters abblack li matters that law enforcent are embedded in the same tworks and doing what they can to sort of morph. I think all in itself and most recent set of uprisings i started to think about back a couple ars to the intercept story about the nypd and ibm and their facial Recognition Technology that was going to enable people to spot suspects based on race or physical facial characteristics, skin color, then to see thats come toruition in places like florida and miami anelsewhere in recent uprisings to identify suspects and people who were involved in these ings. I think all of it leads back to the question ofhose technology are we going to be able to use . D what is at our disposal if we are thinking about technology being marshalein this way . Does a platform like facebook or twitter allow us to do that . I would say probably not or at least not the same way today as it might have for a moment in 2014 or 2015 2016. But then t question is, what is the alternative . I think thats the interesting precedent that joanne raises with this idea of decentralization and choice for a group to say, i want to assemble my technology infrastructu based on my own directions and interest in where we want that to go. I think its an interesting proposal and then my question is really all about the realism of being able to do that at scale to make a significant difference. I think the question of scale is such a hard one. I see it frustratina lot of groups. Its funny because i think about my first group was very critical and ultimaty got some rponses in two levels of response were, this is to bleak and the second that its o u. S. Focus, like open up a litt bit. I try very hard in the robotics book you lots of robotics to expand the frame and it was helpful to hopinbecause i think some of at civic tech in taiwan is real interesting i did a presentation there a couple years ago and they are looking at for example some of the ai plans put out by the japanese governments and i think on the regulatory front the australian approach, they really he a lot more interaction between academics and the governments in terms of research and the governments in rms of thinking about Going Forward and how to maybeuild or go their own way. So i think even though the u. S. May feel its a t of a stalemate right now in many fferent ways, thats probably putting it mildly aspossible. I do think there are other International Examples i really draw some hope from. I think europe with some respect to the datarotection issues as well but yes. Frank, when youtalk about scale ive always bn very moved by you and others who argue that to treat ople as people ithe right way we need to treat them as individuals. Then the question is, when there are so many decisions to make about the many people, that possible with or without technology . Yes. One of the things im tryingo do with my book as im trying to ke it so thati that ranks, sorts, and evaluates people is more expensiveand slower to run. [laughter] i know that sounds that may sound like a modernday ludde approach but its a service of two things, one is to encoure that human touch, partilarly with Sarah Roberts work on content moderaon, take a good depth look, the whole center for crital internet ab theres people doing really importt roles that are really undervalued. To say this is all that of the ai eventually, i think its actually these are jobs of the future professional roles of thfuture. I would say the same thing about the rights we forgotten your. It involves very difficult decisions to say this website when someone searches your name this website should it, but think its gonna be a job the future ideally the replication of juice online. I think what i would like to see less jobs are in areas like security, policing, military arms races so a big vision for my book is to say lets have an industrial policy of ai where we make certain things we dont like more expensive to do and rerouting most people to say ai for healthcare, cancer recognition,ai for drug velopment. All the extent that the should not be pushed to Human Services but pushed away from human control or gargling with the rules. Im told and at 1 50 pm, we have about two minutes left. I thought it was abif any of you have a couple closing words maybe thoughts about things we caaspire to and work toward in the future, if you have an opportunity to comup now would be a greattime. I would love to add to franks point cause this in my mind is to say that this technology does not alwa work. Its bas on a hoax. What is called through capitalism is very often not its far from reading our minds and in fact delivering formation none of us want to see. How many of us here have bought a mattss online and received advertisemen for mattresses for the next six months as people by mattresses daily. This technology as big as it is imperfection is something to continue to stress because it might seem like there might be a very easy way to attack it but i think they are incompetent and incompetence in management, incompetent and delivering anything to our users is its pretty plain to see. I might as well just offer some other addons to frank his int to slowing down both slowing down and asking different questions. When we are looking at designing new systems to begin that with not just what uld help me do this more efficiently as a comny or what have you, think about reframing that question of what helps me do someing that is good . How do i build a housin staging system or mortgage system that makes it possible for more people to get access to the benefits of homeownershiand then designing something with that question in mind. I think all of those things are predicated on slowing down to think about and ask dierent kinds of questions and to the point th was brought up earlier to be able to see and understand the potential harms ofhat it is we might design rather than rushing to simply geing the thing done and getting the thing out. I agree. I just want to thank you christina for moderating and its wonderful to be on the panel with both of you. Now more from this years virtual brooklyn book festival,