ep 32 - On the decentralized web, truth, and human rights: 'Hacking authenticity'

24 Dec 2023 · 15:04

Jonathan Dotan, founding director of The Starling Lab for Data Integrity at Stanford and USC, joins a16z crypto's Robert Hackett to discuss how cryptographic tools can establish ground truth in conflict zones. Recorded at a16z crypto's Founders Summit, the conversation covers the decentralized web, human rights verification, and the history of open source regulation shaping how web3 tools reach journalists and historians.

Speakers

Topics

Transcript

[Music] welcome to web 3 with A6 and z a show about building the next generation of the internet from the team at A6 andz crypto this show is for anyone whether company leader or other entrepreneur Creator or developer Media or policy maker seeking to understand and go deeper on all things blockchains crypto and web 3 we're back with all new episodes beginning with some conversations from our recent second annual Founders Summit including today's episode which covers human rights such as recent crises in various regions around the world and the role of Technology such as a decentralized web our special guest is Jonathan doton Tech founder Emy nominated producer and

writer who spent six seasons on HBO's show Silicon Valley doton is also the founding director of the Starling lab for data integrity at Stanford which prototypes tools and principles to bring historians legal experts and journalists into the new era of web 3 so here he's in conversation with our very own Robert hackit who has joined to host the show a number of times so you'll recognize his voice and who was also a former senior editor at Fortune among other things they discuss how cryptographic Technologies can help establish ground tooth in Conflict zones the history of Open Source regulation and more as a reminder none of the following should be taken as business legal tax or investment advice please see A6 inc.com

disclosures for more important information including a link to a list of our investments hi everybody I'm Robert Hackett I'm an editor here at a16z crypto I got to say while you were just talking about how the government is viewing this stuff I'm getting an intense bout of deja vu like didn't we fight these battles in the 90s about cryptography being Munitions and there was the Clipper chip and I thought we all agreed that it was not the way to regulate this stuff yeah it's truly strange and if you go and talk to folks like David Cham or even Christine Peterson you know Heroes that have kind of established like what we might think of as our understanding of modor cryptography or open- Source technology I think they're all baffled as to what's

been going on and and I must say this is something that is incredibly recent and fast moving I think the consensus around open- Source technology in the last three weeks has evaporated as we've known it we can't rely on the types of consensus that we had before so yes what do you mean by that elaborate on that point well I mean if you look back at the first crypto wars in the early 1990s like we we started to establish like the basis of what we've all come to rely on which is that code is a form of expression it cannot be regulated by the government in all ways and that this is very important for us to sit and take stock of these discussions because with the illusion of this form of imminent threat from Ai and it's not to say that there aren't concerns around AI but the

question is like what are the measures that you're going to take and it's hard to believe that Supreme Court precedents from 30 years years ago are suddenly being ignored and tossed aside so we're in a critical moment to reclaim all of this discussion I have to ask given the work that you do at Starling and your focus on human rights if you had to pick between AI alignment and effective accelerationism what would you pick well let me share with you what I'm most concerned about I think we need to take the tone of this conversation in a dramatically different place and even the question around like alignment my question is not about whether or not you're going to create Harper metrics around that but it's this no no that somehow a form of Applied statistics is going to create in the near term some form of sentient out of control form of

automation we have tons of mechanisms to continue to control technology and I think that if we need to start from that basis there are a series of things that we can do and we don't necessarily have to change laws and we can be compliant with the laws that exist on the books without having to basically say everything is new diplomatic answer so I want to return to some of your work but but first I've got to ask you know you gave us a highlevel description of your career trajectory but I feel like I missed a few of the connecting Dots here you went from writing jokes for TV to archiving testimonials from Witnesses and survivors of genocide I mean how do you go from from point A to point B there well I I think the earliest Parts

in your career are often the most formative and within about a 2-year period I went from on my way to college my dad said can you please pleas please trust me take a job at my friend's startup and work as a programmer just in case all your crazy ideas don't work out maybe you know here's your backup plan uh so in the first web bubble I actually worked in technology I did user interface design and front-end programming and then I started a career in Hollywood afterwards I wanted to follow my passions and that took me all over the world working in 30 countries and then ended up working for the United Nations as well and all of that you know within a very short period of time really created the Ence of of the things I care about so now with something like Starling you know I can I can dip into

those experiences and so you can sit there with International lawyers because I worked with them earlier in my career and and you often think when you do these types of crazy things in your early 20s when am I ever going to use this type of thing again and in the end it all kind of makes sense just takes a little beat so by the way we will open this up to questions in a little bit but first i'm going to ask you a few more questions we have conflicts erupting globally right now I understand you're Israeli and it makes me wonder given that that country was formed in the wake of the Holocaust what sort of personal motivations you have for this kind of work in human rights so obviously the last month has been incredibly challenging for anybody who wants peace

it is a somewhat hopeless situation right now my own motivation on this with the work well first I should say that we are engaged in the conflict and for operational security matters can't talk about exactly what we are doing but what I can say is that we are very concerned about the way in which things have progressed there both in the cause to go to war and also the methods of fighting in the war and our North Star comes from work that I did with the last surviving prosecutor from nerenberg got a chance to interview Ben Fen this incredible human being at the right young age of 100 years old I got a chance to talk amazing when was that two years ago wow and what he told me which is what I

carry forward into this work was that the lessons from nurenberg and in mind you he was the very first prosecutor had ever used the term genocide in a court of law he said that the lessons from nberg are that nobody is above the law and the eyes of the world are watching so take that as you will but that is somebody who saw the very worst of mankind successfully prosecuted all of that in a court of law so those precedents are important to uphold and and they are apolitical uh and they speak to the importance of of how even in the worst times you there is still there are still rules that is extremely powerful how might starling's technology apply in a conflict Zone like Israel or

elsewhere so again I I want to be careful that sure to not be very specific because as a academic institution we have to take even extra measures to work with organizations that do work in the field we we can't be in the field ourselves but we work with partners that are so a couple things I can say broadly first of all that as we look at the information landscape right now there's obviously a lot of challenges with understanding like what is real and interestingly I should say that with generative AI what we have seen is not a flood of Miss and disinformation coming from there generally speaking if generative AI has been using it's been used as a form of expression of protest of creating these hyper objects as it were for expression and I think that's important so it shows that you know we're not in this

immediate situation in which we actually cannot disambiguate but that's that's a very narrow window right now in which you know the tools that are available right now to the masses haven't been able to create the type of confusion that they may very well so we are working with journalists we are working with historians we're working with lawyers and we're trying to find ways in which we can do something very simple which is simply establish the original pixels the time the date and the place in which a photograph was taken that's it and I know that sounds relatively easy but obviously it's a incredibly complicated in those types of chaotic environments and the idea behind this is to be able to establish a form of hygiene around the information and persistence because you have to understand that none of this if it is

ever brought to a court of law will be something that's dealt with in the near term we're looking at something you know upwards of you decades when I was in Bosnia working on war crimes in the beginning of my career I was dealing with evidence that was at that point 15 years old so that's the scale and and the challenge essentially the temporal challenge that we're facing Rafi anadol the artist said something yesterday that really struck a chord with me he said data he believes is a form of memory I feel like the work that you're doing is just like the best possible example of that we've got just a couple minutes left I want to squeeze in a couple questions if we can does anybody have anything they might like to ask thanks

for the work you're doing have you seen any of this new evidence with the tech technology that you're using actually used in any cases in the international court of justice yet and if not what's your expectation what that all happen so the founder of our our legal program at the Starling lab is the current ambassador at large for Global criminal justice and so with Ambassador vak we've been exploring a variety of ways of making sure that folks over at the doj and elsewhere are aware of this type of technology and the Article 15 submission that we did to the ICC was our proactive way of showing the court and briefing the court or I should say the prosecutors at various points to say hey this this type of stuff really really

matters and I think that they've come around to understand this is a possibility and the Hops to let's say analogies here in the domestic or system they're not as far as you might think there is a whole body of Law and evidence practice around self-authenticating information CCTV evidence that can be brought forward from let's say like the New York Times there are ways in which the courts just understand that they're you don't have to go and have the the foreperson of the printing press of the New York Times in court to be able to explain that this is a legitimate front page of the New York Times right so it's those types of traditions that we're trying to draw in to ensure that things can become self-authenticating but I want to explain the Highwire act on this because this is actually really difficult is that for as much as we are excited about

the possibilities of providing attestations of this kind remember there's a very thin line in which this form of authentication of information could suddenly become become a form of surveillance if this is something that a user does not actually know that they're participating in then you could have a situation where a government or let's say a manufacturer or a nefarious agent could ensure that this was the default and they could ensure that basically it's a way of tracking everything back to let's say a specific mobile phone and that could be its own human rights problem and so that's part of the challenges that that we've been looking at and I must say that as abstract as it was originally I had researchers go and look at cell phones out of North Korea and we worked on this standard with the Doby if you have been reading about the

content authenticity initiative they did something very similar to what has been put forward with a form of cryptographic attestations around the Integrity of images but what the North Koreans did mind you eight years before we did is they ensured that the hashes that were created there became a form of content moderation and censorship so there was an allow list and if I sent you a photograph and it wasn't on the allow list then it would be instantly deleted on this forked version of Android that they had so it could be traced back to to me it would also be deleted if I sent it to anywhere both of those things are massive existential problems so that's that's why we have to be careful with this type of Technology quite orwellian I believe there's a book in the library

that we have here seeing like a state which talks all about this sort of like information legibility and how when a state has such information is able to identify and name things it has power over it and then you're subject to the motivations of that that state and a place like North Korea might not beid do we have any other questions in the remaining time thank you so it it seems like ultimately the authenticity of the images comes down to trust in the manufactures of the cameras or phones that took the images so to what degree does that play a role here and ultimately how that would be used as evidence so there's a cleavage right now in the technology path so the professional grade cameras are actually the ones that have the least amount of Technology around this front whereas

mobile phones have excellent cryptographic Technologies there are already trusted execution environments you have a way to access it directly through the operating system and so one of the things that Dan and I have been working on is trying to advise the camera manufacturers of ways that they might be able to create what is currently state-of-the-art for mobile phones with authentication of essentially the operating systems there to ensure that the right OS is installed that same type of system could actually be put into cameras and we actually it took years but we got a kind of concept camera built by Canon this Summer that did exactly this but remember all of this though still requires a lot of work because it's vulnerable and if it's not something that is opted into by the photographer

then it is a form of surveillance so we need to be very careful with this type of technology I have so many more questions that I want to ask you we are unfortunately out of time but I'm going to ask you one more before we exit the stage which is if you were writing Silicon Valley in 20124 give me the one sentence log line what would be the plot or what would happen given current such a world in which we live I mean seeing Richard and Gavin in an MMA fight would be that is great I love that that's fantastic that'd be my vote so I going we to capture it all with starling's metadata thank you for listening to web 3 with A6 andz you can find show notes with links to resources books or papers

discussed transcripts and more at as6 crypto.com this episode was produced and edited by sonal Choy that's me the episode was technically edited by our audio editor Justin golden credit also to moonot design for the art and all thanks to support from as6 andz crypto to follow more of our work and get updates resources from us and from others be sure to subscribe to our web 3 Weekly Newsletter you can find it on our website at as6 crypto.com thank you for listening and for subscribing let's go [Music]