 Good afternoon wonderful to see you all I am Celeste Watkins Hayes the Joan and Sanford Wildein of the Gerald R. Ford School of Public Policy here at the University of Michigan and I'm delighted to welcome all of you this evening to our policy talks at the Ford School event featuring former New York City Mayor Bill de Blasio welcome today's event is presented in partnership with the science technology and public policy program housed here in the Ford School with the support from the U of M urban technology program at Tobin College and our media partners at Detroit Public Television and PBS books. Bill de Blasio served as the 109th mayor of New York City from 2014 to 2021. He guided the country's largest city through the pandemic and among his many other accomplishments he created a groundbreaking early childhood education initiative which became a nationwide model financed the finance the construction and preservation of more affordable homes than any other New York City administration and drove the city council to pass a New York City green deal. We green new deal make sure I say that green new deal. We are happy to have him this afternoon to look at smart cities during his administration the city rolled out a number of smart city solutions and tech education initiatives in every borough earning New York the best smart city accolade in 2016 and recent months he's also become increasingly concerned about the rise of AI and other emerging technologies and their impacts. We look forward to hearing about his continued observations and engagement with the topic. Mr. de Blasio will be in conversation with Professor Shobita Parthasarathi Ford school professor of public policy and director of the science technology and public policy program. Her research focuses on the comparative and international politics and policy related to science and technology and specifically how to develop innovation and innovation policy to better achieve public interest and social justice goals. There will be time for questions at the end. So please scan the question card qr code to submit your questions throughout tonight's event. Those tuning in virtually can tweet your questions to hashtag policy talks and those of us in this room who don't have access to the qr code. I bet you can tweet there as well hashtag policy talks. I have two undergraduate students here to help us with facilitating the Q and A. Audrey Miello from the Ford School and Enzo Magano from urban technology. With that please join me in welcoming Shobita Parthasarathi and Bill de Blasio. Hi Bill how are you? Hi Shobita how are you? I'm good you know it's a very auspicious day and so I wanted to kind of scramble our agenda a little bit. I know we had a sort of discussion about the discussion but as many of you may know today just like an hour ago I was watching this on YouTube President Biden issued an executive order on AI. Some of you may have heard about that and some people I mean it is for the U.S. the most comprehensive AI strategy at the national level that we have we don't have any overarching regulation and as the dean said we are certainly very interested and you are very concerned about questions of AI and for those of you who haven't looked at exactly what it is that that this strategy does I encourage you to do so if you're interested in questions around AI but it includes things like requiring companies that are going to develop foundation models i.e. generative AI to submit information about that to the federal government and in particular to submit the requirements of the red team that is any kinds of potential problems with the technology it includes guidance on how to address concerns around bias and discrimination you know there are provisions around privacy around potentially misinformation in terms of content and watermarking that content it sounds fabulous so but what do you think but you left a hanging there it sounds fabulous but will it be fabulous look right I do want to give President Biden a lot of credit because you know a lot of times folks who are concerned about technology concerned about artificial intelligence you know make parallels to the reality of the debate and the public policy actions around climate change like here are two vastly important topics both with existential ramification and we can safely say vis-a-vis climate change in my view you know the Biden administration has done amazing work particularly through the inflation reduction act on AI this is a good start because having an AI executive order is a major major step for the United States to say you know I am laying down some law some structure some order I commend him the but is two points one as Shelby I've pointed out you know how are you going to enforce requiring the companies effectively to share their secrets now I think it's exactly the right thing to do I think their secrets their proprietary information has a lot to do with our futures and they are making decisions for us without consulting us as if they are their own government that's not acceptable so I think it's exactly right for the federal government to say whoa whoa whoa you're going to have to share this information with us including the things that you see that might be going wrong the red team analysis and that that information has national security ramifications as the executive order makes clear in addition to the impact on our lives day-to-day but the enforceability and what's the mechanism going to be what's the consequence is going to be and I didn't see so far yet in the executive order on a clear illustration of consequence because I think we can all say about corporate America that if there is no consequence good luck getting them to comply and if there's a modest consequence corporate America treats that as cost of doing business and they violate the law take the hit move on with their business so the consequence question is huge and executive order I suspect inherently has limits on what it can do consequence wise and the real action is in the legislative front and part of why this conversation is so important and everything I want to thank you for all you do and your colleagues do in science technology and public policy because we have to have a very different conversation in this country because we have to actually get to legislation if we expect to rein in some of the problems of artificial intelligence and have coherent approaches to the future that can only happen with legislation the European Union is way way way far ahead of us I'm glad they are but we need to catch up so good that this happened today question mark on what it actually will mean and no substitute for the real thing which is for our Congress to actually pass laws related AI so we started out a little bit wonky there we were so long usually that happens in about half the audience already I know they're like technology I think it's nap time no so one of the things you talked about there was existential you know consequences and I'm wondering maybe if you can talk about this I mean we build this as an event about smart cities and now we've started out you know at the you know one million mile level and so I think it's worth at least starting there and then eventually I want to get I want to get to the consequences in the local level as well but what are from your perspective when you think about AI what are the benefits why are you worried how did that worry evolve for you and look I want to we will definitely talk about cities and we'll talk about you know the smart cities concept but I do appreciate that the title of this talks was you know our smart cities really smart you know the reality is I think this really gets to the the AI question by definition there are things artificial intelligence can do that will help us all there that can there are such things I like to use health care as an obvious example I don't have a question in my mind I'm not a technologist I'm not a health care expert I have a question in my mind that there will be ways in which AI will improve health care but at the same time the reason I got more and more deeply involved in the discussion is I've looked at this combination of factors I looked at the absolute and total lack of democracy in decision making unlike the vast majority of things that we all care about where we can identify someone somewhere in public life someone elected someone accountable has to take some responsibility for the decision right now the vast majority of decisions related to artificial intelligence are happening in big tech headquarters with no reference point to all of us they're not asking us they're not consulting us they are not going to wash in for permission they're not going anywhere for permission that for something of this magnitude that is going to reach every corner of our life the sheer absence of democracy a debate of any kind of oversight of any kind of checks and balances alone should scare us half to death in my humble opinion but then let's talk about what this will do AI unquestionably is baking in biases that already exist in our society because it's taking in the exact same reality that govern the status quo in our country and in our world it's feeding it those facts into its systems that the folks who are feeding it into the AI databases and are replicating our status quo and reinforcing it that's a problem AI unquestionably is going to play a role in the displacement of initially millions and ultimately tens of millions of workers in America alone for when you talk about the rest of world hundreds of millions over a very short period time no comparison in history as to how quickly this is going to move will those workers be consulted at this rate no will they be compensated will they be retrained will they have any guarantees today no and that's going to affect also every corner of America and then the more damning and ultimate dangers a technology that can easily fall into the wrong hands I mean we've already seen just with our current technological reality the extraordinary acts of hacking that have disabled major parts of government in the private sector imagine as AI advances what it could be done with it in the wrong hands imagine what happens the level of malfunction because all technology and machinery malfunctions but what a malfunction of a very sophisticated AI system can mean and then of course ultimately the giant unanswered question but the fact that it's unanswered is all you need to know could this technology become self-aware and come up with its own choices that could affect the fate of humanity well the technologist developing AI are the first to say in many cases yes that's really a possibility and then they continue developing it so I mean everyone in this room is smart you know if I said to you hey I'm over here developing something you know here in the room that's really interesting but it might poison everyone in the room you always say hey let's stop that right now and find out how to get the interesting part out but not the poisoning everyone in the room part that's not happening with AI that's why I'm a little worried a little so I'm wondering you sort of talked about this but you know a lot of the people in the room I mean this is an event happening in a policy school and and you know I think some of the stuff that we've talked about is that these kinds of conversations tend to be happening and in our geographic kind of context on North Campus that is in the engineering school in the computer science department and not necessarily in a policy school and in part because I think for a lot of the folks who are here we focus on questions around social policy or security policy or housing criminal justice environment and I'm wondering if there's a particular case that you as a non-technologist right is there some a case or something that you saw in this set of questions that opened your eyes about the connections you know which you were just drawing between AI and technology and these other areas of policy yeah I mean obviously the worker piece you know I ran for mayor on a platform of addressing inequality and a lot of that was economic inequality and to me you know I tried to use as you heard affordable housing and pre-k for all and a number of tools to try and create some greater parity some greater possibility for economic fairness well when you juxtapose that against a technological change that could cause massive job displacement it kind of takes us backward potentially and that was extremely upsetting to me because there it would be one thing and it's your I like your two-campus analogy I ran on the tale of two cities in terms of income disparity and also you're talking about tale of two campuses your tale of two campuses plays out nationally the folks making decisions are extremely economically secure if you want to come with me to Silicon Valley and we can ask people there you know each of the ones making the ultimate decisions about their net worth I think we can put them in the category of making more than the rest of us and so they're fine but they're making decisions for a lot of people who are barely scraping by so right there I'm worried but then the notion do I trust and I'm being very plain about this do I do I trust executives in Silicon Valley to think about the truck driver in Michigan and whether that person is going to have a livelihood once an autonomous vehicle takes over their route I don't have any reason to trust that tech executive that they are even could think what the life of that truck driver is like and that's a tale of two countries really that's a tale of two Americas that I don't I imagine the tech executive thinks they're doing something good for the world because it's right to say there are efficiencies and there are things that AI will do for us that will better certain parts of life I agree with that what is shocking to me is to think how do you miss the part about all the pain that could come with it and I do think it's a it's a very pertinent American problem and it's well documented that sort of our our common experience has dissipated you know not so long ago in America most people did not get a college education and not so long ago in America people were much more commonly from pretty humble origins now we have this massive skew where again that truck driver you know they may well feel that there's not a plan B in their life whereas the tech executive you would cannot even imagine ever being without resources and a livelihood so I don't want the tech executive making decisions for any of us I certainly don't want the tech executive making decisions for someone they cannot possibly understand nor do I think they particularly care about and again when you're talking to the tune of millions of jobs there's 13 million who take every kind of truck driver in America is 13 million truck drivers and autonomous vehicles are one of the most obvious examples of how AI and automation are going to displace employment now so let's say you say oh no it's only going to be a quarter of the industry okay two million three million people this place in one industry and it won't be slow so what are the solutions from your perspective for example particularly to this question but other questions around AI and its potential impacts what are the kinds of things that you've seen or the ideas that you've seen that you think that that hold promise I think it is amazing and I'm not trying to say what the European Union is doing is perfect but I think it's amazing that you know by the end of this year at least it's projected they will pass legislation that has a lot of teeth some of the same things in President Biden's executive order but much more extensive and much more teeth and I'm trying to remember the exact percentage I think it's 6% forgive me if I'm wrong their penalty for certain violations of the law the European Union proposed law the penalty for corporations that violate the law is 6% of global earnings so not the penalties in the thousands or tens of thousands these are penalties in the billions now that's how you get someone's attention you know that's how you get a big company's attention so I was super impressed at that kind of model you know and it takes 27 nations in consensus in unity in the European Union to agree to anything that's kind of stunning that that's happening I mean while we are not even the first base in terms of legislation in Washington so I'm glad that you want to talk a little bit about the framework of that of the AI that's the simplest way I can say it is some of the same points that that should be talked about with the the Biden executive order that the companies have to and you'll certainly add whatever I missed the companies have to divulge their work they have to be accountable for problems that they find in their work that the there's ideas of liability in terms of the impact of the products that are being created it's very wide-ranging and I think the way to think about it is that it it kind of I don't want to say it reverses the situation that wouldn't be fair but it creates a dynamic where the reckless development of AI is penalized and it creates guardrails in my view where in you again get in the corporate mindset get in the profit-making mindset now they have real reasons to think much more carefully about what they're doing and to feel that they are directly and specifically accountable so I think that's the sort of world of difference I see in what's been put together there but I think the solution in this country since we all understand that there's a challenge getting anything done in Washington right now the solution is state and local actions whether it be legislation executive orders whatever might be done through state and local governments and by the way institutions every major institution including universities have to come up with their own policies governing the use of AI and every time they put some kind of limitations some kind of transparency some kind of oversight in place that helps the overall trajectory and the last thing to say is activism now look I want to just offer this in the spirit of recognizing that everyone in the room has a voice and everyone in the room can make a difference there is no singular activist movement that I can identify in the world right now related to artificial intelligence there are activist movements that touch artificial intelligence and I think one most obvious examples is around privacy and addressing surveillance government surveillance police surveillance certainly there are activists working on that issue I'm not trying to diminish that certainly not trying to diminish the folks in the labor movement who are working to ensure that displaced workers are protected but there is not if you I've tried this if you go try and look up a movement that is about artificial intelligence and what it will do to all of us and and ensuring the public's voice in that debate I don't think you'll have an easy time finding anything of any scale if I said you go look up movements and organizations around climate you know the list would be long if I said go look up regarding the rights of women and and whatever side you want to say of the abortion issue the list would be long so how is it that on an issue that will touch all of our lives every single one has massive unknowns and dangers and is guaranteed to displace a huge number of workers by the admission of the people creating it they're creating it to displace millions of workers it's not a mystery how is there not a social movement I think the answer is there needs to be a social movement because social movements and activism ultimately drive policy more than any other factor and I think people including in this room can be part of building that that's super interesting and I I want to get back to the social movement question in a minute in part because I think the language of disruption in Silicon Valley is at least 20 years old if not longer right so we've been talking about disruption for a long time but not but there's sort of assumptions around that about the positive the fact that everybody will benefit and there are no risks I think we need to disrupt their disruption so I want to get back to that in a minute but I'm wondering if you can talk a little bit more about the state and local level so policy today right so social movements often take a long time to build long time to produce change but what are things that states and localities can do now are there states and localities that are doing passing policies that you think are a step in the right direction I mean look I'm happy in New York and I'll be I'll be transparent about happy unhappy I'm happy in New York that we took initial actions to recognize the use of algorithms by government and to try to address some of the ramification of that obviously particularly in terms of bias I'm happy we passed legislation that that the first one happened in my administration the second legislation the post-act to address police surveillance and to create some guardrails and some transparency around that process step in the right direction since I left office legislation that puts real limitations on the use of AI in the hiring process you know these are steps in the right direction unquestionably and I think any time a state or local government did that kind of thing it is starting to create consequence starting to create limitation which is what we need but I think we're only scratching the surface I look back my confession today is you know I wish I understood better I wish I had prioritized more of these issues when I took office almost 10 years ago because I now see how hugely this will affect our lives and that we've got to you know write down to what we're spending our money on and what we're demanding of these companies state and local government and institutions need to use their financial power their economic power to start to put limitation into the equation we've got to start grappling with this question around employment we didn't we didn't feel it a lot in the time I was mayor the displacement in terms of employment but we could see it over the horizon I wish we had moved more aggressively then but there's still time to do it for example you know what states and localities could do is say here are the rules related to when a worker is displaced by automation and artificial intelligence you know here's how they get compensated here's how they get retrained and they could put that obligation on a company we in New York in previous situations have played past worker displacement legislation that requires companies to account for the workers and think about that for a moment and I want to blow your collective minds here you know imagine a world in which we said a working person and their job is more important than the profit of the company that displaces them what have we said that is a societal value we're not telling the company they can't create their new product we're not telling them they can't profit we're saying the old ground rule that the working person is simply cannon fodder or collateral damage in technological progress we don't accept that ground really more we actually want that worker to be protected in an ever more unstable world because by the way if you could ever have argued that technological progress and its its negative elements were okay because of the thriving American middle class I mean who believes there's a thriving American middle class anymore there's tremendous economic insecurity and so I think it's a great time to revalue things and say as a matter of policy if the worker is not protected then there needs to be a consequence of that company that's a very different way of thinking but states and localities have started down that road and the more that do it the more possible it becomes for others so if you don't mind my disclosing something you said earlier today to the whole how dare you to the internet oh my god one of the things that you said that you you know that it that the Detroit Lions would win the Super Bowl yes I admit it I admit that I said it I wish you kept it to us the pandering we absolutely appreciate the pandering is you know what you said this morning was that when you stepped into the mayoralty that the tech companies came calling right that they were there telling you you know singing you sweet lullabies about the fantastic future that the investments would bring in tech right and I'm wondering you know you you are obviously painting a slightly more dystopic picture but to get to that place to see the tech in more color you need capacity you need different kinds of people at the table you need different kinds of processes and different kinds of policies so what what can cashstrap cities do to see you know to take off the rose colored glasses when they when those are the when from their perspective the old the only way they know how to see is through those rose colored glasses it's a great question look I I think the first of all to the to the opening yes that you know suddenly all these tech executives want to be my friend and it was not because we shared values it was because you know they saw dollar signs and if they could get New York City to be a customer or to regulate in a way that they liked which actually meant not regulating at all and it quickly became apparent that we had very very different views of the world and I had my tangles with Uber and Airbnb and others but I think the point that you're really good underlying point is for so much of America what we've been told for so many city-states everyone we've been told you know more technology equals more efficiency equals more effective government equals lower cost for the taxpayers and sometimes that's true I don't even mind at all saying yeah when that is the case that's cool but there's another side of the rainbow which is when the price of admission means destroying your regulatory frameworks around things that were actually meant to protect people in communities or protect working people or provide revenue for the government so the government could provide services I mean there's all I mean basically the story around a lot of these companies is trying to evade you know health and safety rules regulation rules related to workers safety and working conditions wages and not paying their taxes like that was their model that they thought was a great business model and you know the rest of the business community even folks that I had issues with were paying their taxes were abiding by all these other rules right so this is where what started as a friendly conversation soured pretty damn quickly but I think for a lot of America what we've got to do is say of course we'll take the efficiencies no one's trying to throw out the baby with the bathwater but don't take the poison pill you know don't take the other things are gonna come back to bite you and and this is where I actually I'm an eternal optimist I really am and I was mayor of God forsaken New York City so I'm an optimist after that I love my city but like I like to say I governed over nine million highly opinionated people and the most diverse place on earth and like the strangest things would happen every day and I'm still an optimist I do think this kind of one day will not be a red or blue issue I think one day it will transcend because what's going to happen is first of all liberty and privacy are valued very strongly on left and right for all of you want just a kind of beautiful example the Patriot Act which is passed after 9-11 and then was later on renewed and had an underlying value but also in some ways a lot of civil libertarians felt went too far the alteration of the Patriot Act to kind of tone it down was actually a left right co-production in the US Senate where Rand Paul like co-lest with some of the most left-wing senators to force changes because they both were coming from their version of a civil libertarian worldview well I think on this one I mean that makes that's child's play compared to what AI is going to do vis-a-vis privacy so I think there's a left right possibility there I think there's another left right possibility of power and I hate to say this but the displacement of millions of workers is going to be equal opportunity red state blue state rural and urban I think there's going to be a cause there for people across the spectrum to say wait a minute this changes the equation and to demand accountability and compensation of some form from the companies this kind of thing where I could see even more conservative leaders and jurisdictions saying we're not going to just take this wholesale and one thing you just said about efficiency the role of efficiency I was I was thinking it just sort of popped into my head as you were talking but you know for example in the policy school we teach students about public management right that's one of the required courses is public management and there was a time and I think to some degree it's still there that you know the focus is on efficiency better practices etc which kind of leads us to technological solutions but in fact I think in the Ford school and in many places around campus but I'll just say specifically I know the Ford school a little bit better so you know I know at the Ford school we're training people to go and understand for example you know criminal justice housing what are the real problems that are happening to people that get lost when these decisions become algorithmic right so that's a way in which it's actually what I think implicitly it sounds like what you're saying is we need to reintroduce the experts the specialty experts right the people who understand housing or criminal justice or foster care as opposed to assuming that the technology is the expert or the technologist is the expert in the process of bringing in algorithms we're destabilizing for example the sorry to say jobs of a lot of you all too right I mean that's part of what's going on and I presume you know there's that kind of dynamic and what you sort of started to see probably you know given the timing in New York the early stages of that perhaps yeah I mean like how about humanity makes a comeback you know like bring the human voice and the human perception back into these decisions because no the algorithms cannot pick up on an immense amount of human reality and nuance and they have incredible baked in biases by definition and this is again the false idol of efficiency I'm not saying there aren't examples that that again don't work as you can find some examples where you can have efficiency that's actually just efficiency without bias and without negativity but what I don't like is how undiscerning it is it's like oh we've got a tech solution oh it must be good you know if I said that to you about anything else in the world you'd be like what the hell you talked about like you know like show me the thing before I tell you it's good right but we have been we're I think this is a fair explanation of the American psyche going back a good century or more we've all been told you know progress equals technology and it's you know just let it flow and it was never true it was never a pure truth it's become even less of a pure truth and and the other thing to say and I use this analogy you know previous technology came with oversight I I kind of love repeating this and I don't think I said it yet in this session I said it earlier nuclear power nuclear weapons which you know if anyone says oh AI is so complicated like that's why we can't get in the way of big tech or we can't regulate them or we can't oversee them right you know because it's so complicated only they understand well I assure you the people who created nuclear weapons you know that was complicated it wasn't like everyone could just figure it out but they knew and it was very public they knew it needed to be very very tightly controlled by the government and then you say oh those are weapons but what you know that doesn't count well nuclear power the same exact thing nuclear power from the very beginning had extremely tight regulatory schema and when nuclear power started to cause immense problems there was a lot of public activism there was a lot of legislative action and then there was real decisions made to reduce the role of nuclear power in our country now no one in the night I saw some of that play out no one said oh civilians you don't belong in this discussion because you can't explain nuclear power to me so I can't explain the nuances of AI but I damn well belong in discussion because it's going to affect me that's democracy so there's an attempt right now and I put it at the feet of Silicon Valley but I think there's many many enablers in all sectors of our society there's an attempt to say this should be treated unlike anything else in our history and it's technology therefore it's good but in this case it's technology and it's so good you should have nothing to say about it and that's where we should all be freaking out because no one's ever told us to butt out before in such an overt and aggressive manner and when I don't know about the rest of you someone tells me you don't belong in this discussion you're new you don't belong in this room you don't belong at this table that's the table I need to be at so let's talk about disrupting the disruptors a little bit so what is your tell us about your vision for a you know global AI social movement what would that need to look like what would it who would who should it include what should it do I mean I don't I don't want to pretend to invent the wheel here I mean just take the climate movement as a real easy parallel you know I mean it should be something felt in places like this for sure you know a lot of the stuff by definition college campuses and progressive local jurisdictions often like the match for this kind of thing but the notion of having a people's movement that puts a simple set of demands on the table I certainly think the notion that companies have to be held responsible for the products they create we would say that about any other product we say about everything automotive right if you have something's wrong someone's going to get sued and everyone accepts that liability well if AI leads to something that's harmful to humans shouldn't that liability come right back on those same companies that's something we can demand we can demand that workers who are displaced you know the companies involved whether it be the you know the companies in the particular industry or the companies that create the AI that displaced the worker or both have a responsibility to that worker and their family that's a that's a financial material equation that's something we can put on the table we can demand a certain level of transparency and democratic decision-making we can demand that our congress pass real legislation we can demand the states and localities pass the kind of legislation I mentioned before that we did in New York it's not perfect but it's a very important start there's plenty to demand my the reason I'm agitated besides being a New Yorker the reason that this is how we normally talk is that the reason I'm agitated is like none of that demand is on the table right now in the public debate now the good news is we could have that you know by next year this is not a it's so complex that we can't build some of this quickly if people get involved but we've got to start having those focal points of opposition I'll use an example which I think is a really good one the fossil fuel divestment movement I'm very proud to say that New York City in my administration divested five billion dollars of our pension fund money from fossil fuel investments and that got the attention of some oil and gas companies for sure you know divestment has been a classic example I'm sure there's a variation of that you could apply to AI and big tech but the point is people people took it upon themselves to say okay what what do we have in the venue we're in do we have the power to invest differently do we have the power to you know get contracts from and procurement differently do we have the power to demand different limitations I think that kind of movement can be started quickly and then as we saw with the climate movement it can be knitted together around the world right now you would say the climate movement is absolutely international and making very coherent demands in common across many nations that's a great model to apply to the AI situation where does so we've seen in recent even just weeks and days I mean today right we've got the writer's guild that incorporated provisions around AI I mean fought about AI incorporated provisions into the new contract the auto workers are doing the same thing where do you see the those both you know in terms of this budding potential social movement what role might they play but also what do you think about the ways in which they negotiated and the kinds of things that they achieved I think it's fantastic I mean the fact that you know the writers obviously this is particularly a discussion about Hollywood when it comes to the writers and the actors but the writers really got some significant guarantees about their work not being displaced by AI is it perfect no was it really important that that's a step towards making this the norm also the actors are still not back and I think it's really important you know this this country kind of runs through its streaming services at this point you know the fact that you know the people we all love to watch are on strike right now and on strike in part because of AI and they're taking a very very tough stance is very encouraging to me so I think the labor movement is actually you know the this has been an amazing year for the rebirth of the American labor movement I think it's going to grow but I think a lot of it is the way AI is coming into the picture and actually galvanizing working people to feel that there's a threat they must confront so what so we've you sort of talked about the role of universities generally and I I know you are not a university employee by any to my knowledge no I am in a university oh are you doing damage I'll be back why would you think I wasn't what do you say about me I'm a visiting fellow I see that as a as a cop as a cop visiting fellow at New York University okay but they haven't finished my background check when they do they won't they won't keep me so so I mean then you that means you're really on the spot I was gonna compliment your ignorance but you know you know there's a reason I want to say that's that wasn't rude I it may sound you're like what is she just say I I said earlier in a room when you were present I said I have very strong opinions and sometimes our opinions on issues I don't know all the nuances of but I my quote is I'm unburdened by knowledge helps me get to very strong views go on so what what can universities do what can when students leave this room what can they do what what are the steps what do where do you think universities can play a role in this I just think no one should underestimate their voice and I don't mean that to be you know a beautiful romantic statement I mean it because I've been involved in the political process my whole life and there is no such thing as a voice that can't have an impact you know if you if you're consistent enough and persistent enough and you join with others so just with social media alone I mean if you're worried about the issue start talking about it if you heard something here today you thought was meaningful talk about it the if you think this is an important issue form an organization or or if one of the organizations you're in now can start to do something on this issue pull them toward it as I said I think you know what this university obviously is a huge university what role can it play in creating more accountability what role can it play in putting demands on the tech sector if tech folks come to this campus start asking them these tough questions they feel that you know I've seen the trajectory with some of the masters of the universe here like who were very arrogant a couple years ago and they're getting more and more getting their asses kicked as they should with people saying what are you guys actually doing like what are you doing to us and that's helping because they need to feel accountability too so I think you know there's so many prisms for action you and you know you're also residents of the city you're residents of the state what can the city do what can the state do I think the important thing is to be begin with the notion that just talk about it just say to people what you feel just say what solutions are starting to make sense to you talk about the impact on working people because especially the extremely rich history in the state of Michigan in terms of the rights of workers like this and and because again all things automotive and particularly a trucking industry are going to feel the brunt of AI infused automation this should be one of the places where there's the most agitation and and you know one of the things someone said to me a long time ago which I really appreciate if you you know if you don't see the social movement for the issue you care about go start it so one of the things that I sometimes hear especially among you know sort of in in the policy school is but I don't understand the technology right I'm I'm not a technologist I don't have an engineering degree or a science degree how can I you know have any confidence that I understand the issues how do I engage you know I I'm not the one who belongs other people belong and of course as you said yourself that's something that the tech industry takes advantage of in various ways and I'm wondering you know of course you are in a privileged position you were in a privileged position certainly when you were mayor and and still now but how did you handle that when I presumed people pushed back and said well you don't understand and I'm sure you see that now well you don't really understand what's going on so how not only how do you manage it but what guidance might you have for others who are in that situation well I just think it's always ratification of truth when people tell you to stop talking you know when they tell you don't belong when they tell you your ideas impossible whatever I always assume that's like the guarantee of the opposite so you know do I need to understand how to code to be able to see what the impact is going to be on a worker no or do I need to understand you know their nuances of their technology to know that I'm not into discussion or none of us are in the discussion no that's very very plain to see and I just don't accept the idea that expertise is what gives you citizenship you know I mean because think about any topic you want to talk about foreign affairs or the military or transportation or you name it how many people in this room whether you know as I took a poll of each room how many of you are experts in foreign affairs and yet every voter every time there's a national election is in effect voting on questions of foreign affairs and national security that none of us are pretending like I didn't go to West Point I still have a right to an opinion like that's that's actually baked into our founding documents that it is not a nation of technocrats and or decision-making class of technocrats and experts it's actually a respect for the everyday person that their lived experience is what matters most and here we have lived experience on steroids because it's going to hit us from all these different angles and already even before sort of the more intensive application of AI I mean just on privacy alone just on data alone people are pissed off they're they're perfectly smart about this point they're like you're using my private information to make money or you're taking information from my kids without our consent you know or you're feeding information to my kids that are dangerous to them so you can make money or any permutation this is well understood by average Americans so actually like I'm like I trust the average American to get what's really going on much more than the people telling me that we don't belong at the table now I want to hear from these people too I agree so why don't we start with some questions come on let's go hi so thank you so much again for being here we really appreciate it and so the first question submitted before the presentation started was during your time as mayor of New York City you launched the link NYC program to provide free Wi-Fi and internet access to residents building on this experience how do you believe the United States can expand similar initiatives nationwide while addressing concerns related to data privacy and security well I mean I think link NYC was a mixed bag first of all to be honest so it's kind of one of those things that noble experiment with some good and bad elements I think it's though I think if I understood the question right it goes right back to the notion of localities have an important role even that one that helped shape national and international policy and the thing I would say that's really this is my lived experience now speaking as a former mayor the to say there's no international governance is an understatement you know there's an occasional agreement or treaty that has some impact but there ain't no world government and then our national government you know kind of flickers on and off in terms of its ability to deal with the issues that affect us all and there are days when our national government seems to be a full government doing full things and there are many days when it's shut down or on the verge of shutdown or there's no speaker of the house or whatever it is and this is not just this year we've seen variations on this now for the last decade or two why do I say that because in the end the only places that actually have to govern all day long are states and localities like we don't get the shutdown especially localities and that's the irony the the the responsibility levels you would think that the higher up you go the food chain the more exalted and responsible people are the least responsible form of government is global because it doesn't really have any shape or impact then you go to national government here and in many other countries that are essentially paralyzed a lot of time you go to state governments that at least by and large take some responsibility but the place where all day every day responsibility is being taken in some kind of solutions are being attempted as local government and so I think that's how to recognize even on an issue of this vastness go where you know who's got the hot hand right the local governments the place where you can get something done at least this question was also submitted before the talk today is often said that technology can be a great enabler but also a divider how can we guarantee inclusivity in the advancement of smart city technology ensuring that no one is excluded or is it inevitable to leave a minority behind in the pursuit of progress oh wow no it's not inevitable it's not progress if you're leaving people behind and that's I think beautifully summarizes everything we're talking about right now we have a form of quote unquote progress that is inherently built on leaving a lot of people behind I think the way to see it is to turn that equation around and say it's not progress if there isn't a democratic process it's not progress if there isn't transparency it's not progress if there isn't a debate and to recognize that we do not have to accept things this way I mean I think that's the thing that really I want to drive home we are being told this is the only status quo there cannot be another history tells us that's a falsehood there is no such thing as only one status quo there's no such thing as an unmovable object in terms of a democratic society so we have to say no if your discussion leaves out a lot of people it's not a real discussion now we're moving on to questions from people who are present or watching the event live so our first question is from Anjali and it kind of dovetails with our prior conversation about having a seat at the table and making sure your voice is heard so they want to know oftentimes even when tech companies are more transparent relative to their peers about what they're developing the takeaway is still so jargon filled or difficult for a layman to understand how do you think cities can start translating across the language barrier to create change I love that I love that yeah there are people who use language you can't understand on purpose you know you know many people come from families with different backgrounds and the classic example of you know if the kid doesn't speak the language from the home country and the parents want to say something about the kid you know they say it in their native language that the kid doesn't understand well that's like what we're dealing with here the tech industry is using a language they know we don't understand on purpose if they wanted to make this conversation more accessible if they wanted to bring us in there's a lot of ways to do it I do agree it's up to government to do the translation and to empower people by saying we're gonna we're going to take these concepts and make them much plainer much more simple for people you that does not mean dumbing down you know but but trying to portray the ramification because I don't want to hear about the the glorious theory or the elegance of the formulas I want to hear what's it going to do to human beings and that we can put into plain language here's another question submitted many of the perils of AI come from its place within a high capitalistic system the loss of job is important but not because of its inherent value of the job itself but rather what the worker can access with their income with stronger safety nets I universal health care education or even the idea of minimum monthly income create the conditions in which we could take advantage of these technologies to capture more efficiency in the business in the business world without the threat of worker survival I would say a forceful yes and no there there is a absolute grain of truth in that theory that so much what I'm saying is about the insecurity people live with all the time and if you address that core insecurity you'd be having a different discussion about the impact of technology fair statement yes someone call me when we have universal health care and you know those kind of income supports and and proper education for our children you know the day we get to that if you said we're suspending all development of AI until we get to that that would be a very interesting conversation obviously what's happening is AI is racing forward and social progress is not in fact we're going backwards sadly on education because of the pandemic housing costs are going up for so many people not down so we're we're totally out of balance the other thing I would say is a lot of people I think the no part is a lot of people try and use that kind of formulation to say you know this is just about the economics when in fact I'm saying there's a democracy problem here too if if you took away worker displacement out of this equation all workers did great but still these decisions were being made without us and our privacy was being invaded and the technology was being developed in ways that could be dangerous to our human future I'd still have the feelings I had and I'd still say unless there's this thing is subject to real transparency in democratic debate I ain't buying it all right so you also spoke a lot about how there isn't a coherent or cohesive social movement around AI right now and perhaps there should be so the next question asks if the answer is a social movement around AI how do you get those negatively impacted so the truck driver in Michigan for instance to care about and act on these issues well I would you know we just saw some truck drivers take the country to school with the UPS strike that didn't have to be a strike you know I mean that was as I as a progressive and someone who believes in the labor movement I'm watching the labor movement right now labor movement got game I didn't used to and now it does I mean there was a great there was a Wall Street Journal headline yesterday that said the UAW you know keeps the big three off balance again you know because they're just there they're so on the move and they're so agile and obviously the Teamsters with UPS and in a different way the writers with the Hollywood studios so no I think I think there's so much to be said for the notion that working people can assert themselves effectively in this situation I've lost track of the original question because I got so agitated so the original question was how do we get those negatively impacted to care about and act on these points thank you my point it's one of the things I learned in public life is when you don't know what the hell's going on just say it and so yes leadership tips where am I what exactly is this gathering my point is I think working people are already asserting themselves have you talking about here in the state of Michigan obviously with UPS being a great example they're already asserting themselves or the UAW being a great example and I think the notion that they can assert themselves a lot more is it is live now in a way I might not have said five or ten years ago for totally different reasons than AI for a lot of other disruptions there's a moment of movement in terms of workers demanding their rights now can they coalesce with other movements I always believe that like-minded people can find a way to each other you know that that common enemy and I hate to use that phrase but it's just needs to be used in this case that the person who cares about privacy can find common ground with the worker who cares about displacement can find common ground with the person who worries about democracy being undermined and I think again one of the things I encourage people to do is you know if you feel some of these things that we're talking about here go talk to people who share those views including people who share those views but have another motivation because I use that example of the Patriot Act I mean you'd be amazed when you start talking to someone that you think is different but you know you have something in common with how much that one commonality can drive action and the only way to find out is to try so you mentioned in your talk that this AI talk this AI discussion is very similar to the climate discussion if AI is bigger than climate the risk of course is that AI becomes politicized in the same way as climate and eventually actions on the topic is deadlock how can we avoid this it's a great question but I would say I'm a climate not only am I not a climate denier I am a climate change action believer I think the trajectory and look it was a tough stretch when the previous president took us out of the Paris Agreement and and time was lost and all that but there is a there is a not unfair or overly optimistic view that says look a lot of countries have started to move more than they were a lot of the private sector has started to move inflation reduction act was huge and surprising in its scope and that all comes from the climate movement and it all comes from the climate movement's ability to work at the local level and bubble up the solutions and create demand so I would say I'd love to see the AI movement mirror climate movement will there be will there be politics around it of course will there be divisions of course but I actually think the climate movement is broadly bridging right now there's fewer and fewer deniers thanks unfortunately to hurricanes mudslides wildfires you know all the extreme weather is kind of wiping out the denialism but the climate movement is showing the ability to make a lot of impact even if it has a high number of detractors out there so I think that's a good model for an AI movement so speaking a little more generally you I wants to know if you could wave a magic wand what are some smart technologies or climate solutions that you'd like to see implemented today in every major city well look I think there are elements of both AI and you know renewable energy that could go together and really be helpful I mean if you had to simplify how do we overcome the climate crisis it would be you know a really really aggressive transition to renewables and you know there are issues around renewable energy even like the transmission of the energy the storage of the energy or how to make how to help people do their own part with the renewable with solar and all and link it to larger networks you know that I do think something like AI plays a role in addressing so again I like I have these benevolent moments where like oh look at this puppy AI and it's going to do nice things for us right and then it turns into like you know a Rottweiler and so it's like the good part can be applied in a way that I think could help us move forward so long as all those other things the democracy the checks and balances etc are in place in relation to AI with the revival of the Office of Technology Assessment in US Congress which was shuttered during the contract with America movement in the 1999 era be effective and effectively regulating AI is there a way to translate this to work on the local level yes there's a way to do it on the local level but we had this conversation in one of the earlier meetings it should not be an office of technology assessment off to the side somewhere issuing an occasional report that someone might or might not read you need very central figures like cabinet officer cabinet official level or in the case of a city a deputy mayor you have to be really serious now about the kind of weight that would be given to unofficial and their office that's assessing what the hell is really going on with these technologies what they're going to do to us and I think it goes beyond the traditional sense of assessment it has to have an sort of action focus because the real-time dynamics now I mean it's kind of what back when there was an office of technology assessment it was kind of quaint what the assumptions were about how the the conveyor belt of new ideas to action to impact on humans would be and how much time we had to sort these things out we do not have time anymore stuff's moving real fast and has to be real-time decision-making and you know folks have said for example should there be rather than attempt that legislation alone should there be like an FDA for AI or something like that that you know a bunch of experts that could make decisions and I think it's actually a very productive idea except the FDA takes for frickin ever and we don't have that in the case of AI so it has to be more centrally located in the government and more supercharged in terms of timing NYC supported tech hubs that were focused on bringing in underrepresented minorities into this into the tech sector and expanding computer science education for every student do you feel that these were successful in meeting these goals who are all these smart question ask I'm like wow what show me that what is this place it's magical yes computer science for all has had a huge impact and with all my tech angst I 100% believe this is the way of the future to retrain a lot of teachers and how to teach computer science skills and make it much more broadly experienced subject in public schools and also ensure that the tech sector is more representative which only happens if you reach across the spectrum of the public schools that like the tech talent pipeline and the other attempts at employment good I don't think great but good I think we could go a lot farther because again now on the happy side of tech you know a lot of a very high percentage of tech jobs can be reached with a an associate's degree you know and obviously some require a higher but there's there's a lot of opportunity that still is not being offered to a more broad constituency so I think that was directionally right I don't think the results were stellar obviously the industries are going through evolution now but I think it's directionally right to say how do you look at the way that industries are staffed and not just leave it the way it is but try and create some kind of governmental presence to ensure both broader opportunity and broader diversity in these industries I was submitted this question you mentioned that over time different party interests can converge to make real change such as the renewal of the Patriot Act how can we ensure that the how can we share that such a convergence of interest happens with respect to AI before it's too late and millions lose their job without a safety net well look I this is this is what I worry about all the time and think about all the time and I think you know again by all the things I've said the raising our voices and acting locally and acting in our institutions and building a social movement but I think this point about before people lose their jobs is a really important one we are not on that trajectory to stop the first wave of massive job loss and I hate saying that but I can count I will say that is going to change American politics profoundly and I my kind of evocative way of saying it is regardless of what you think of President Trump if if you think Trumpism was a response to economic and social dislocation and cultural dislocation if you make that analysis and say Trumpism was the result the residue whatever way you want to say it of rapid social change that was positively like glacial social change compared to what we're about to see so I'm saying millions of people will lose their jobs in the course of just years not decades years and that is going to reshuffle the deck of American politics very rapidly now I don't like that's not what I wish for I just think it's what's going to happen and when it happens I do think that will open the door for a different kind of coalition a different kind of activism and action because you cannot have you know if millions of people get dislocated that means their families beyond them which means now you're starting to talk about is it 10 million people is it 15 million people what is it it's a hell of a lot of people it's going to cause people to question and it either is going to be regressive in ways we might have said was true in the last few years or it's going to be an opportunity for a new kind of coalition for change or some rich mixture but it's going to be a moment for action in any scenario okay so the idea of just portraying the harms of AI to constituents feels like it could potentially create a tech dystopia that while more warranted than that while more warranted than the current AI AI hype is still falling on the pendulum of tech extremes do you think it is possible or even worth it to try and communicate both the pros and cons of the technology in every day to everybody who are these highly intelligent people nuanced they're so nuanced it's a great question and I will always tip my cap to some balance I am never going to say it's just one thing AI but I think we are so out of alignment in the public debate that I would rather take a more militant view because it's fact-based there is no democratic process around AI literally doesn't exist there's no accountability there's no transparency decisions are being made about us without us every hour of every day and look I'm not a technologist when I say there's existential threat up to the level of the machines becoming self-aware and taking actions against us that's not because I saw a Terminator 2 with Arnold Schwarzenegger it's because the technologists who developed this technology acknowledge their own fears about that again parallel to nuclear weapons and go see the movie Oppenheimer which is extremely powerful and moving and the fact that the people making the nuclear weapons acknowledged in real time their fears but then they went and acted on their fears and immediately tried to limit the production and the proliferation of nuclear weapons very energetically like the people who created them put their backs into trying to stop this from becoming the bigger problem it could be and by the way as of this hour they succeeded so where is the parallel in the tech industry of people saying you know what we've we've there are things that are good they're coming out of this but there's a hell of a lot of problems here and we're not just going to go and say oh government please regulate we're actually going to energetically fight for change and put some limits on ourselves and have this discussion in a way that's honest about the dangers that's not happening so I think it is incumbent upon us citizens to say this is broken and if you want to accuse me of some dystopia you can accuse me of that but it's a hell of a lot better than a techno optimism that leads us on the road to hell you know I'm like I let me be a little extreme to help be one of the people balances the equation because the other team right now has all the points on the board and we better even up the score a bit for in favor of the citizenry ask me how I really feel he's like how do you really feel I like that how do you feel about the current nyc administration's changes to your past policies to regulate technology or make technology more accessible to residents nice try I don't get into that whole discussion I look I think every administration is going to have a different approach and every administration is working with current knowledge and I'm not inside the you know the building hearing all the latest I do think if any jurisdiction attempts to be sort of tech friendly they need to also ask the tough questions about the dangers and be ready if they don't like what they see to move quickly so again I don't mind I don't mind any sort of the traditionalism of we would like tech to work out for all of us that's kind of I get it just yes as Ronald Reagan let you say I don't have a Gerald Ford quote for you but as Ronald Reagan used to say trust but verify so we've spoken extensively about the potential future harms of AI but someone in the audience wants to know what we're doing to address the more immediate harms of AI that are already here so you've talked about autonomous vehicles and California just revoked cruises autonomous taxi license license because they kept hitting people so how should we be addressing these kinds of AI downsides right now just shoot me I like because they kept hitting people like this was a movie you knew what was going to happen before it was over right it was like come on let's let's create a car without a human in it and assume it won't hit people I'm like come on you know like this is not mysterious so I'm sorry what was the question what we what we should be doing to address those immediate harms right now I think California did the right thing after a long time not doing the right thing I had a team of very earnest technocrats come to me years ago in city hall and say hey we've got this great idea for autonomous vehicles and part of Brooklyn I was like no you don't you know it's like no I'm like no no no all places in Brooklyn that the cars would get beaten up I'm worried for the cars like are you worried for the pedestrian no I'm worried for the cars so no I think the right now limitations the worker displacement piece is a right now thing that we had as I mentioned we had New York City worker displacement legislation that we passed in certain industries that delineated the company's responsibility when they changed the job and they put someone out what they owed that worker that's something to be done in state and local level all over the country right now vis-a-vis the impacts of AI the transparency dynamics again there's so many institutions and localities that have big contracts that are meaningful to the tech companies they can start demanding more transparency in the process in exchange or they can threaten to pull their contracts you know the this is kind of the oldest story in you know in terms of the political process like you know there's that great phrase also a very New York phrase it's not about the money it's about the money and you know which is kind of follow any political discussion and it will often lead you there and so purchasing power and economic power can be brought to bear starting right now because I think these companies need to understand that there'll be a consequence and the consequence can come from the consumers it can come from the institutions so there's many many things you could do but the point is I'm just riffing a few there's a whole host of right now actions I'd like to see that demand I'd like to see people start to issue that demand and I don't care who those people are in the sense of I don't care if you're a college professor or if you are you own a business or you work in a business or you're a student or you're a city council member I'd like to see people start to issue the demand so in addition to AI another thing that makes a lot of tech smart cities is the use of cameras and you mentioned in your talk that people's personal information is at risk with this new AI technology New York police department spent a tremendous amount of money on surveillance technology like ShotSpotter what changes would you make and is there technology that you feel is beneficial yeah I think ShotSpotter is beneficial so ShotSpotter to me and to bet look from my vantage point I'm happy to be educated otherwise but I authorized it and I did not see civil liberties or a privacy problem because it became an incredible tool for knowing that you know it basically tells you when a gun has gone off and speeds the response to it also what I did not know in my beloved city was a lot of times a gun went off and we didn't know anything about it because no one got hit turns out a lot of people who carry guns are not particularly good shots and I'm really happy about that fact so no there was all sorts of gunfire that never got reported to the police because there wasn't a person with a bullet in them but actually knowing where the gunfire was and being able to recover the evidence of it often related to how you stopped other crimes from happening or caught criminals for other reasons so ShotSpotter to me was a was a virtue the bigger surveillance question which I think is very sensitive one very real one clearly you know New York has a particular concern about terrorism and all these has since 9 11 that's not changing and it shouldn't change in the sense of we are you know in many ways the number one global target we have the UN you know the capital of American finance everything else but one of the things that happened with the post act for example coming out of city council was to say the police have to have like really stringent rules about how they monitor the use of technology they have to be able to sort of publish their approaches they have to be able to answer questions about anything that might be used inappropriately that create a notion of you start with even if you have to use this technology in certain circumstances it's not a blank check and it should be altered over time if you find excesses or mistakes and I think at least we have a framework for that so like this is a classic example we have a framework in New York City by law that says these things are not already written like there's no 10 commandments of how to use AI that you can't violate in fact it's the other way around right now surveillance technology is being used we want to constantly see the updated ground rules we want the ability to question them we want to know of it for some reason if there's actions that are not consistent with our values or consistent with the stated objectives and we want the ability to change when we see something needs to be changed that's the law I'd like to see that be the law everywhere and those were the last of our questions thank you so much thank you well done wonderful wonderful thank you so much well done god perfect timing what an incredibly smart erudite incredibly good-looking audience wow we we already invited you to campus yeah that's right now that's right that's right it happened all right thank you everyone I hope you found it helpful and appreciate very much excellent conversation thank you thank you very much