 Okay buddy. Hello all and welcome to the Hyperledger Media Entertainment Special Interest Group. Today is March the 4th, Monday. Morning Ethan. This event is being recorded and will be available on the Hyperledger YouTube channel. We'll post that link on the, it'll be on the wiki as well as we'll post it on the LinkedIn Hyperledger Media and Entertainment page. I am Brett Russell, Chair of the Hyperledger Media and Entertainment Special Interest Group and with us again today is our fabulous and very helpful Chair Assistant, is Randy Givens. Good morning and welcome to Hyperledger Entertainment and Media Special Interest Group. Randy will provide links to the chat to the Hyperledger Code of Conduct, the Hyperledger Anti-Trust Policy and she'll put in the links to our LinkedIn page in the chat so please open those up and bookmark them and join the page, follow the LinkedIn page and you'll be up to date, kept up to date on everything that we're doing here. Yes as you can see the links have already posted and you're welcome to click in and to join our community. Thanks Randy and Randy will also be monitoring the chat for any questions and comments that may come in. Thank you very much everyone for attending. Thank you Brett and please leave any comments or questions in the chat. Thank you. Excellent, thanks Randy. So as I said earlier, I am the Chair of the Hyperledger Media and Entertainment Special Interest Group and today we're going to move on to our second Blockchain AI Roundtable. We are sure a few important members today who had other important commitments. Todd Holmes who's the Todd is the Associate Professor of Entertainment Media Management at California State University Northridge, Jay Devrant, Jay Devrant, Jay is the Manager of Deloitte Consulting, Blockchain and Digital Assets and based in Toronto and we had some great contribution last meeting from Andy Rosen who's the Technical Innovator and Troubleshooter. Andy is the Founder of Sequence Key. Today we have some very important people and Orson Wimms. Orson is the Executive Director of the Music Education Initiative and the nephew of the legendary Al Bell, Founder of Stax Records and past President of Motown Records Group. Ethan Kuhl. Ethan is a Senior Data Specialist with Wal-Mart and also works with us with Karen Kilroy through the non-profit Friends of Justin. Karen Kilroy. Karen is our moderator and a CEO of File Baby, an author of a number of great books, Blockchain, Tethered AI, AI and the Law and Blockchain as a Service. I am Brett Russell again, Founder of Accuradius which is a we're developing Blockchain in the entertainment industry. I'm also the Chair of the Hyperledger Media and Entertainment, especially this group and the organizer of this Blockchain AI Roundtable. Today we have an agenda, a special agenda that's been put out to share amongst the members here in our group over the last week, two weeks and we've already had a lot, some very interesting discourse on this and I'm excited that that we got some fantastic feedback about the current efforts and this is in specific to number three. In fact, I apologize. I'm going to put the agenda up here in the chat and oh I'm sorry. I could do it if you don't have it handy Brett. I have it right here. Thank you Karen. Actually go ahead. Can you throw that up there? I apologize. I got it. Well Karen's doing that. I'll just run through the four items that were slated, are slated for discussion today and if the clock runs out, we're going to continue the discussion online at our next meeting whenever that might be and any other way we. So the first agenda item is should the entertainment industry create a category for any film, TV, music, streaming, production that used AI during development, industry, benefit or detriment? I wish that Todd was here today because I think he'd be the first guy I'd hit that up with. Hey Karen, he definitely is the go-to guy for that. The second item is blockchain permission and public ledgers. They have time stamping. We can store data sets, rights metadata, training data, content identifiers, digital fingerprinting, monetization, smart contracts. I mean these are all things that that AI is data and this works. Blockchain works with AI. There's no doubt about it. Standards groups and agencies should participate. What standards groups and agencies should participate in policy making for the use of AI in the entertainment industry? Again, I know Karen has got a lot to say about this and I know there's some groups out there that are involved but it's big business right and the final one is training data that's used to produce new music, film and TV, video, script writing, etc. should be protected, regulated, licensed. So there's a lot to chew on here and I think I don't want to use the numbering that we put in here and start off with number three, the standards groups and agencies and give the floor to Karen here who's got a lot to say about this and has an absolute wealth of knowledge. So Karen, the floor is yours. Well, thank you, Brett. As Brett indicated, we've been up to our eyeballs in content, provenance and authenticity and he also mentioned that I had written a book, Blockchain Tethered AI, which came out a year ago on Valentine's Day and Blockchain Tethered AI was the product of years of research, about seven years of research went into that on how AI could be made more trusted with Blockchain. Meanwhile, there are many other standards groups also who had their eye on AI and other fake media and trying to figure out ways to prove authenticity there that were not involving Blockchain. And so there's really, really a lot of good ideas from what I would call that side of the house, right? It's almost like two different worlds for us because we've been in the Blockchain world and meanwhile other people are coming from industry verticals and they're addressing these problems. So it's almost like the old Reese's Peanut Butter commercial, Reese's Peanut Butter cups, I got my chocolate in your peanut butter, oh, you got your peanut butter in my chocolate. So that's kind of where we need to be right now. I think from the technical side of the house, especially the enterprise Blockchain side of the house, which is quite a thing in its own merit. There's a lot there, a lot of the problems that the industry vertical groups are bringing up as problems that we already have the answers for. So there needs to be a lot of meaning in the minds. And one of the things that we've been involved with on today's panel, we happen to have Orson Weems, who is president of File Baby, and Ethan Keele, who has helped me come up with the idea for File Baby. And he's a distinguished architect at File Baby. And what we have built is based on a practical applications of standards that were built by a standard that were developed by the C2PA standards group. And so C2PA stands for content authenticity, or I'm sorry, coalition for content authenticity, content provenance and authenticity. Someday I'll get that right. And then there's also another group, CAI, which is content authenticity initiative, which is a broader group where people apply the technologies that are developed by the C2PA. And so since today's panelists are, since today's roundtable is Orson and Ethan primarily, I'd like to really dive into this topic and into what is C2PA? What's interesting about it to us? How did we get into it and why? And I'd like to start with Orson. Orson, can you tell us what C2PA is? Yeah, it's a standards group because we need some type of a way for people to know that we're not just pulling rules or standards out of the in there and there's a support system behind that to make it very comfortable and for it to be something honorable for really to have the type of responsible responsibilities and regulations set. We can't just have things running a mock out here with this and we need to have some standards set. And what we know is that the major corporations and the media companies that are joining the C2PA, some of the majors that we're part of through Friends of Justin, a nonprofit that we're part of so that we could be involved with that so that we can let people know that we're part of a standards group that really is interested in the responsible AI usage and the delivery of that AI so that people know that when we tell them that it's something to authenticate and to provide that provenance, it needs to be backed up by something and we have all of these organizations that are behind us and it's very important to know that trust is a big part of this because some people have seen some of the horror stories already as quickly as AI is moving and the uses of AI as moving. So I think that we need to have this and it's important for groups like the C2PA for us to have these type of standards that we can go to and see some of the violations that we're going to see coming out later on and it's important for us to do that. Yes and Ethan would you like to talk for a minute about how we got started with C2PA? For sure yeah that was a great explanation Orson thank you. So about this time last year maybe a month no it's been about 11 months probably we were I was at the Fayetteville library with Karen as she was doing a series of talks on how do artists begin to integrate with generative AI and learn how to make money with it? How does it improve their creative process? How does it improve the business side of their art? And as we were going through this Karen was showing off Adobe's Firefly which is an image generator and I noticed that there was a little icon in the top little information icon whenever an image was generated and lo and behold that that little icon was a C2PA icon and basically what it told us was hey this image was generated by AI through Adobe Firefly and I thought that that was pretty revolutionary for a company to already be integrating AI like AI watermarking or AI manifest to let people know but this is actually generated by you know AI and as we explored this more and got involved with C2PA we joined C2PA in August the use cases really just started to blow up and we really began to realize the power of someone human AI everything in between that being able to claim and basically put your name on your work on the internet and be able to claim its fair use is going to be crucially important for the future of the economy and for humans in general we really need to be able to tell what was human made versus AI made how it was made in order to really protect human rights in age where AI generated content is you know proliferating and we have to be able to establish a value system between AI generated works and human generated works in order to protect the ingenuity and the the humanness of work that we put online and so that is really where it began in and where File Baby was was really born was with this conception of how do we enable people to put their name on their work and have it stored safely on the internet so people can see Jane Doe created this this photography using you know a Nikon and then they also use Adobe Firefly on top of that to to you know change some things in the background generate some images in the background things along those lines I think it's really a a crucially important thing to be able to distinguish for the future of the of the human race yeah it seems like this is coming on us it's almost faster than what we can comprehend the the changes are are almost daily and you know anybody can feel free to jump in here but like one of the things like for instance you can't you almost can't promise you can you weren't going to use generative AI and everything because it's it's like you know you think about the automation that's come upon us you know a simple version would be the calculator right the calculator made it so you don't need to you don't need to to do all your math manually anymore and there was a time when that was people were aghast about that right they were just like shocked that you know why would you do something like that that's cheating you'll you'll you won't you'll lose your skills you'll never be able to figure anything out again you'll be stupid but here we are you know quite a few years later and we realize that you know of course we're going to use new tools as they come along human races is progressing really really rapidly and one of the things that I saw that is a case in point really for I think it was a kind of fits with your number number one here is that the that the BBC came out with a statement on how they're going to use generative AI and this was just the other day and the BBC did not say they're not using generative AI they said they're going to be transparent about where they use generative AI and and they they they released a really honorable statement a very thorough thorough statement that I hope becomes like a almost like a paradigm for the industry that people start doing that saying you know you we're not going to be in denial but what we're going to do is be transparent and BBC was one of the leaders in developing C2PA and they still are very active and and so what that means is that they well they started five years ago because they wanted to be able to prove when something was fake news or not you know they have a reporter that's been out there on the front lines they've seen what actually happens and then they see what goes around on the internet and so it's really really critical to a news organization like the BBC to be able to get right back to the origin of that media and be able to track it down so so that's that's kind of been you know what's happening so let me go to Ethan again what what are your thoughts on on how the industry how industry can I just ask a quick question of you just going back to that could you have a link to that BBC statement can you put that I wasn't planning on it but could you put that up in the in the LinkedIn page not not now not sometimes today it's moral whenever you find it I think that's very very interesting and I think that that if that's the start of these standards that we talk about that that's really valuable I'd like to pursue further discussions and kind of follow that as part of our as part of our group here sorry to interrupt you go ahead Ethan sure so I think Karen what you were saying is how can industries begin to integrate C2PA into their daily workflows if I was understanding your question correctly yeah well I'm more like if if you're not the BBC but you're another organization that's tasked with how are we going to use generative AI you know what's a good way to start with your policies and and and your transparency and I just kind of wanted your thoughts on that yeah that's a that's a really it's a really great question so really I feel like the more transparent we are the better and really that is to preserve the integrity of you know communication media everything we've seen you know I read an article I think it was a week ago where I was like how the internet destroyed media in three decades you know something like that and well I don't know if that's fully true I think that generative AI can really have a huge impact on what what can we actually believe and what is the integrity of what we are communicating either to customers to our co-workers um or just you know between businesses um so for me I really think that each company really needs I think I think this is a revolution it's happening right now where each company really has to take some time to think through their AI strategy and be willing to communicate openly about how they are using AI especially generative large language models forecasting things along those lines have been around for decades and I think you know that while those models you know are are also hugely important and there have been plenty of case studies of how forecasting can actually be you know very biased and and harmful there is a new paradigm with large language models and being able to actually understand the the English language and how we communicate that we have to really be able to explain how we are integrating this into our business practices and this also goes for image generation and and everything in between and so I really I really think that everyone's AI policy needs to start with a public declaration of this is how we use it and this is how we plan to track it to ensure that we're not building something that goes off the prescribed rails um you know there there is a lot of worry about how AI can you know create a doomsday scenario and while I think we are still a ways off from that um it is important to begin to think through the policies of of how we actually want to keep AI within a regulatory box um and how and I think this is especially relevant to ensuring that humans are are compensated fairly for work that they do whether be with an AI partner or whether it be completely by themselves um that is really my my biggest concern is the economic factor of how do we ensure that there is integrity and that a company isn't just going to you know rely fully on AI um to for its creative endeavors and for its communication style um because if so we we really have to be prepared for how we are going to ensure that everyone can get compensated outside of maybe a traditional job um whether that be creating content for AI to ingest whatever that looks like but there's a huge I think it's a it's a pandora's box honestly uh there are so many questions and we're only going to be able to answer maybe one a year um if that you know uh because it's it's moving so rapidly I just think it's important for companies to to begin to think how do we use this and how are we totally transparent about how we are using it to ensure that customers have um good clarity into their business practices um so long with an answer there um yes sir go ahead no no sorry I thought you were done keep keep going I long yeah long answer go ahead no no no sorry about that no no I was uh you know TLDR like I said I think it's a PR it's a pandora's box uh there is a lot of questions the most important question in my mind is how do you communicate how you're using AI and how do you show AI work um versus uh human AI work versus purely human work Mike I wanted to ask Gorsen oh sorry Karen I'm sorry Brett I wanted to ask Gorsen how that applies to the entertainment industry were you thinking the same thing uh a lot a lot because you know that a lot as we know and we've seen that uh that as I've said in our very first SIG is that this technology has taken and enhanced or taken and misused certain artists works and whether it's music film videos if it's lyrics etc so but what's very interesting is what Ethan mentioned was the the icon that he saw on the adobe firefly of when we were looking at some of this and building some of the things that we built out and what's very interesting is that the content authenticity initiative and that would be um online content authenticity initiative.org and we mentioned c2pa earlier and that's c2pa.org also but what we know is that that we like to have the content credentials that's that uh the part of that manifest that Ethan mentioned that puts the data in it behind this behind what people actually see you can go in and actually see I like to call it the the DNA of that of that particular person that owns that uh content and in the content authenticity initiative put out a had a panel over I think over in India there were some people that were with some of the larger corporations then in the content authenticity initiative that was a part of that presentation where they wanted people wanted to know and the part of the slide was what are content credentials and and if you will there's some bullet points that's very interesting about what content credentials are and content credentials are a new kind of metadata for images audio video and they said streaming and live so that's an incredible bullet point another one was context attached to content context attached to content again very something that's really needed as we find out as we've said what's AI generated what's human generated and if it's human generated how does that human generated piece get to the point where it can be utilized and not just use free all right what content credentials are a vehicle for authentic storytelling through transparency again authenticity and provenance is that we we talk about so much another bullet point was a way for creators companies to achieve attribution for the content they produce and that's very important also another point was a foundation for consumers to make informed trust decisions and finally the last bullet point that those part of this content authenticity what are content credentials is it's a tool for organizations to safeguard their brands so I think that's very important for us to as we move on this and as as we share this link and this but as we show this hyper ledger presentation that we have now that people really need to understand the impact that we're seeing and we're talking about entertainment from the aspect of designers we're talking about the different things that so many people are involved with that don't get mentioned as part of the music media and entertainment so we've seen quite a bit of the things that we want to address and be helpful about it and we need to be as I see us doing right now being proactive to make this happen so that it can help people I think this is what we really need to look at in having the types of things that are available for people to get comfort because I don't think they need to be hesitant about AI it's already here it's already and we know what it's capable of so and we need to be in that position to share the information that like we're doing right now and do it and and trustworthy and putting out these wonderful presentations you all I appreciate the opportunity thank you orson um Brett were you wanting to jump in yeah I wanted yeah and I apologize for cutting you off Ethan you use the word public and you use the word I mean a lot that Orson and Ethan have just stated so eloquently comes around to trust including identity so my question to uh the expert in the the panel or you're all experts but specific to blockchain and pulling this all back towards our item number two in conjunction with C2PA Karen Orson and Ethan where do you see blockchain and the capabilities of blockchain in particular permission to blockchain and its marriage to public blockchain where do you see it fitting inside that and that's my question to you Karen and to Orson and to but Karen this is a crucial component to our goals and our modus operandi to find that uh you know you guys talk you AI experts talk about modeling you talk about all of the the necessity to learn for AI to learn that information should it be protected should it be documented should it be so that I think Karen I think you you got the gist of my question how can we pull all of this back this great information of the direction that C2PA's headed the other groups that you guys have talked about um and including file baby where do we pull all this back into blockchain which can form that level of trust so that everything after that is let's say provable so that's uh that's my question here and thanks for the uh the spot go ahead we were thinking about doing that with friends of Justin starting a trust group and putting on blockchain our own trust group that's an AI trust group and putting it on blockchain and starting that off as an experiment you know to see how well the blockchain tooling fits the needs of an AI trust group uh so uh that's something that Ethan and Orson and I have been throwing around um now to uh answer the bigger question um blockchain is like a uh like a catch all for all these groups and all these standards so you can make one trackable traceable audit trail of how things have gone because if it ever comes to trying to negotiate with someone for money or proving things in court you know what you really need is that trail of evidence here's what happened when we did this we did this we did this now who knows whether whoever you're talking to is going to honor the things that you've done in a in a court they would generally honor some kind of evidence over no evidence so it's a blockchain can help keep everything organized and also keep everyone involved that's supposed to be involved um through through a trust network uh and distributed nodes now um that said um I will say that C2PA is blockchain ready and it it almost likes walks almost all the way up to the blockchain but doesn't do it it it creates a fingerprint of the file to prove uh that the file has not been tampered with and so there's uh so that's very similar to what we do when we put some put something on blockchain we don't usually put the whole file on blockchain we would take a a hash of the blockchain and put the hash on blockchain and so then uh you can tell if the file's been tampered with based on the hash so that immediately you can see it's exactly the same thing you'd put those hashes on blockchain and and then that would give you that extra layer of of distributed nodes and chained blocks I mean like that's what blockchain is is chained blocks and if you don't have blockchain you don't have chained blocks so that's you know a really important thing for making sure something hasn't been tampered with can I ask you one quick question Karen like that statement that statement is it this not just the perfect scenario that statement by the bvc that they are going to commit to doing to transparency would that not be almost like the genesis block to a commitment from them or a commitment network yeah we commit to transparency here's our genesis block you can look back at that in 10 years and you can see our commitment to doing this noted hashed uh up on ipfs wherever but there's the commitment to doing it and and knowing that that commitment is immutable knowing that everybody in the world can see that commitment on the blockchain means that they can't change from that commitment sorry that's I just wanted to make that point about how powerful what you just said is tying that back to your bvc claim or the bvc article about the the commitment to transparency it's one thing to commit to transparency it's another thing to prove that you've committed to transparency go ahead thank you yeah you can prove it and it's an extra layer of proof um I mean you can it's kind of like the the the it ties everything together it's the glue that holds it all together is is what blockchain should be in this scenario and it's just going to take a while for the different standards bodies to realize that they need it the the the thing that's going to help that is there is now a marketplace of companies that specialize in blockchain and ai and it's sprung up over the last couple of years you know we're right it's exactly a great combination because you've got a technology that can go rogue that needs to be tracked and traced and you've got something that can track and trace it I mean it's just it's a perfect match and so so I think what you're going to see is as these standards come into play c2p is the most mature one and widely adopted that I know of but there'll be others and and uh there's going to be a whole marketplace that can say we can put this on blockchain for you all you need is this plug in and it's going to start to become commonplace can I can I just add a c2p a is part of the linux foundation correct no no they're not no no uh no are they not are they not okay okay sorry they're uh they are a product of uh you have to read the website to find exactly how it came about but it's it's uh news agencies and adobe and microsoft okay trying and they started to they started around five years ago to uh disprove fake news because the field reporters would go out and they they'd see exactly what was going on and then they'd come back and say why is everybody distributing this fake thing when I know that isn't what happened and so they they it came out of a need for that and then the ai was kind of like you know happy accident uh if you will you know we have something that can help track and trace the uh the rogue ai so it just all kind of means to come together um so I guess I'd pass that off to orson uh your thoughts on on what blockchain brings to the table um with uh c2p a and file baby you know we can talk about five baby by the way orson is the president of file baby um and uh that is uh the idea file baby was the idea that ethan and I had to implement c2p a tooling because it was too hard you had to be a programmer to do it so you know orson would you like to talk about blockchain and and c2p a and file baby first what we've seen with using and utilizing of file baby the the way that file baby was set up and it's literally a file management system for to protect a creator's authentic creations their original art whether it's music art publishing any type and what we've looked at now some of the research and you mentioned early on Brett the Karen uh wrote the book AI and the law and our discussions also serve as a way to let enterprise groups large groups know that uh an additional step for security even beyond of what we have was set up with file baby for authenticity and provenance is the blockchain connection to where we can really add another layer of security that uh for comfort for these folks uh we did a uh we've had a lot of contact with the since the Munich Security Council meeting early last month uh to know and one of the topics was again election interference and so with what Karen is discussing and Brett what do you know and ethan and all those that may know how the blockchain works is that we these are some of the things that can utilize the blockchain for the added security of what is authentic now whether people want to want to have in their minds made up that what they have seen that's not authentic uh it's up to them but the proof is as with some of the attorneys we may have spoken to that see what we're doing they say no the provenance is the key and the authenticity so blockchain as that other layer of that and once we understand uh or not what we understand but once the the consumer or someone that was subscribed or purchased or allow us to build the platform that they need for us to implement these type of procedures the blockchain uh part of the of the of what we have are discussing today i'm telling you it's very critical that we have this in place so that people really can go back and say no on this particular day in time at whatever it was this is the original message now what has transpired from that uh that's out there in the world but we know here is the original for people to prove and look at and those that need to make the decision the decision makers they can come back and see and say well this is the real deal this is what it was this is what it was meant to be before it was mis manipulated or misused so uh the blockchain part of that is really an added layer and a much needed layer but again we'll go back to the trust of that and we need to have the things in place like we're talking about that people know it's authentic and it's genuine so that's where i think this this blockchain can play a big part into some of these folks that are delving off into it without folks wanting to make a quick buck on AI and throw something out there that looks really pretty and say oh i'm doing this this and this and then it fails but we want to have it for the longevity because we see how quickly this is changing even since our last uh SIG meeting Brett this this has already changed so much and so many articles have come out with talking about what people are creating people are giving you shocking all-in-bills and whistles and look at how i'm generating and lawsuits yeah so it's it's it's to have this kind of discussion that we're having on this hyperledger uh meeting it's it's critical for us to have and i'm looking forward for you to share this and you're ready to share this because we need to be the proactive people along with the folks that we've talked about the C2PA are just vital the CA vital for us to do this and we're glad to be a part of that and to share with folks and we want to help we really do and we want people to realize that their content has a value because we know the large language models will eat them up is what Karen likes to say and they will utilize it so we want to help protect people all right the as Orson was saying the the uh the content has a value and and one of the main reasons what well when i went to uh Stanford to the uh C2PA meeting in in December uh one of the things that was talked about is how uh they've tried training uh AI models with artificially generated data and what they find out is that after about five training cycles and Ethan i'm coming to you next Ethan is a senior data scientist at Walmart in his other life but after about five training cycles the models start to collapse in on themselves they actually destroy themselves they don't work anymore it tears it up so you have to it has a need for human generated data and as Orson said you know one of our biggest goals is to have an impact on poverty by helping people get paid for their training data. Ethan you want to talk about that for a minute? Sure yeah that's a that's a wonderful point um so when we are looking at contrastive learning and and really trying to I'll back up a little bit when we are looking to fine tune and add layers to a model you generally are either fully retraining a model or you are adding you know more layers to it and and keeping the original as it was and with the problem with doing highly generative contrastive learning and fine tuning methods is exactly what Karen said is the model begins to learn things that are not exactly human nature you know there is push and pull of the patterns that we see and that's why you see if there is a model out there that let's say is trained using chat GPT data it's generally a smaller parameter model and uh meaning you know still seven billion parameters but not on the scale of hundreds of billions uh the trillions of parameters parameter tell them what a parameter yeah that is you can think of it as a neuron um in your brain you know each parameter would be like a neuron uh in your in your brain um and basically what it learns is uh the the connection between two things so the you know in forecasting the connection between um the humidity and it raining you know something like that it builds up that relationship and it's like over cycles of training it says hey this is actually a really important piece of information in this given scenario so we'll activate this neuron so that the rest of the network can can um can react and in correspondence and so when we're looking at something that is purely generated by AI really all it's learning is relationships between different things and our AI isn't exactly human so of course it's going to start to learn that patterns that are not human in nature um and especially the more parameters you have the more those connections uh begin to create a huge network effect of why this thing is wildly off now because um it's learning a lot you know there's so many different relationships that are happening it's it's learning both human and AI you know basically and that was kind of what we had in blockchain tethered AI in my book right that was the premise and that's the thing that made me first think of it as we were we were uh build an AI for people who are blind and visually impaired and so we wanted to make sure that I had asked the question of IBM senior engineers this was eight years ago and I said well how do you know that the data hasn't been spoofed and you know if you're giving this to someone who's blind and visually impaired and you're using training data how do you know that that data hasn't been spoofed and they huddled together this was on a big stage at the think conference they huddled together and they came back and they said if you think that's bad think about the algorithms and I was like holy moly so there's all kinds of uses for blockchain all the way down into the guts of these systems to make sure that they're not an opaque box that they are trackable and traceable that we can trust it even though we can't see into it you know the with the consumer facing trust logos that are then can be dug into so you know these things are all really really important so you know even you know how has the system been trained that's you know really really good use for blockchain and then when you tie it back to the things that Orson's been talking about making sure entertainers get paid for the use of their likeness you know now you can see it's all one big circle so it's really important that standards like C2PA be able to talk to things like machine learning operations systems that have then also posted their usage of the training data onto blockchain and and record of whether or not that person has been paid can also be tied in Karen can I ask a question of you and Ethan and Orson our item number four which you just identified as being that's training data that's used to produce music film tv video script ready etc should be protected and regulated and licensed is there the op and you know blockchain aside because blockchain obviously has some value in just that the the the tracking of that training data but in your guy's opinion should it be licensed could you license some training data could you regulate that training data could you find a way and blockchain would be the proof of what that training data is to actually monetize that training data using the blockchain and then I was but one step further and leave it open because we're down to seven minutes using the blockchain a permissioned private blockchain then leapfrogging over to a public blockchain and having the owners of that protected and licensed training data paid via some type of digital currency that's on a public blockchain like Ethereum so instantaneous of that blockchain recognize or something recognizing that training data which was licensed to somebody being used it could be monetized go ahead whoop there it is or so you think that's too hot to handle that's too hot to handle all right tell them why I said those things well well the whoop there it is uh Al bell that was mentioned that I was the chief operating officer for is the legendary Al bell he uh he owns the whoop there it is and he also was one of the writers of too hot to handle so that's why she said that but uh you're just blazing I'm telling you I saw I saw Ethan was just he is so excited I need to go ahead and answer that yeah I mean I think you hit the nail on the head uh 100 you know that's um that is I think the ideal for our interaction and creating an economy around AI um and it's really a circular economy we're we're building a stronger relationship with AI AI is building a stronger relationship with us we don't want AI to train AI we want to be the ones that both benefit economically from from that process but are also we're building relationships you know and as weird as that is that is the the paradigm shift that is happening you know people think of these massive models as almost like a compression of reality you know there they have done so well at learning everything that they can generally understand most things um and we want to be the ones in control of that we don't want them to come up with their own representation of reality you know um and so I really think that there is a circular effect there that we can tap into like you said Brett where we are getting compensated for helping train these models but we're also in in effect also building a stronger relationship with with the AI systems that we're that we're training with our data so I think you totally hit the nail on the head that's yeah that's the that's an outstanding answer Orson and uh Ethan um you guys get it I mean I think we all get it Karen that everybody here gets just what we have to do to make sure that AI doesn't overwhelm the entertainment industry in all in all respects and there's a lot of work to be done and there's go ahead Orson go ahead well great I went with the article came out last week that uh open AI put out the Sora model of that creating image creator generator and it's it's it's mind blowing some of the the details of that I was really stunned I had to really look to make sure I didn't see this puppy rolling in the snow that you know it looks so real and the the spaceman and all these different things but what was very interesting was an article that came out with Tyler Perry and that Tyler Perry was I was about to or in the plans of expanding his studio and production facilities and he has put it on pause because of what he has seen this application do and he is making a lot of business people moguls filmmakers reconsider some of the things that they do for the large expenditures of putting in a sound stage production studios and all these different things but it's something for us to think about because they're still uses for the content creators to do it the right way and I just wanted to mention that for those that didn't know but this made a lot of people think of what's coming with the AI and some of the things that are happening my my response to that I didn't know exactly the article that you were referring to and I I did I think I did something on LinkedIn about it but in my response it was something like I am very familiar with I can't believe it's not butter butter which isn't butter and I know instinctively when I'm eating I can't believe it's not butter butter and I think that the same is going to apply when I'm watching AI generated video or something deep fake whatever that it's I believe I can believe it's not butter butter type of thing so it's you know you instinctively know I mean some people aren't and I think the majority of the population is not going to be arson to your point absolutely it's worrying that that Tyler Perry has scaled back his plans to allow content creation by humans and and step aside and let the content creation by robots take place so so many things on our on our on our agenda talking about should they declare it you know should you have a big sign at the start of releasing this is all AI generated and then half the audience leaves I'm not interested in you know I can't believe it's not butter butter I'm interested in real butter right like anyway that's my outstanding point go ahead here I just will say that when my my little boy was little there was blues clues live at the palace theater okay he was a big blues clues fan knew who blues clues is with Steve and let's find our first clue is a thing for toddlers and he was huge huge blues clues fan he was probably four years old and so I took him to the palace theater to see Cleveland Ohio to see blues clues great seats and it wasn't Steve it was somebody else and I could hear all the other kids talking about it that's not Steve that's not Steve my little boy didn't care that that wasn't Steve he just tuned him all out and he did not care that that was not Steve and he enjoyed himself to the maximum so you know there is a place for he knew that wasn't Steve because he could hear everybody saying it but he still wanted to be entertained listen it's 1 o'clock 101 and I don't mean to shut this down because we could probably go for another hour but I want to I want to thank you Karen and Orson and Ethan you guys and Randy thank you for paying attention and what's going on and filling in those I appreciate everything from you guys and we'll see you next time look for the recording we'll be in touch thank you thank you to all our presenters thank you so much take care everybody bye bye bye bye thanks Brett bye bye bye