Steering Committee - 2026/09/16

Howdy folks. Hello. >> Hey David, how you doing? >> Good. How are y'all? >> Yeah. Hi, Carlen. She just made it back in time. >> I haven't added your last note yet to the meeting minutes because I just had to >> Don't worry about it. I can I can go do that. I got to get it up anyway. >> I I I added to the meeting minutes um because I had a discussion with Thomas this morning because he can't make it to today's meeting. Oh, okay. All right. There's there's that docu there. Oh, >> well, that's some serious notes about that meeting. Okay. >> Right. Okay. >> All right. Oh, why not? I'm going to add it. Jeez, we're going to have really full agenda here. >> I had to add it because it's our meeting reporting >> our both our TAC and our all members meeting and >> Oh, right. Right. >> So, we have to cover it this meeting because next meeting week is like the day before. >> It's like the day before. I know. Exactly. >> So, >> all right. So, it's two minutes. Let's give folks one more minute. In the meantime, I'm going to go see if I can find a New York Times article about this. I've heard bunch of people mention it, but I have not actually seen it myself. Okay. Oh. Um, >> there it is. >> Who has host for this meeting? Can somebody kick out your notes later, please? Oh. Let's see who has host. >> Oh, I Yeah, I can't get host. >> Yeah, who said I'm not host either. Who said is host on the I can go check. >> Oh, I do have a hoski. I can grab this. >> You do? Okay. >> Oh, great. Thanks. Thanks. >> Yep. You're a host now. Excellent. >> All right. I can't get Okay, I'll go back here. Okay. All right. We're getting it before after. Let me see if I can grab Well, I can do that later. Um, let me try that. I'll open it up there. Uh, okay. Um, you want to share your screen, Carlo? You want me to? Who do you want to? >> Um, I I can share the screen. I have a I have a bunch of screens to share so I can share that. >> Okay, why don't you do that because that way I will also grab a few more links for that last agenda item. >> Okay. >> Uh, while I'm doing that. So, cool. All right. I guess we can call the meeting to order. And the antitrust notice, Carly, is at the top of the page if you want to display it there for all of us that have it memorized by heart. We don't have to read the whole thing, right, David? >> No, no, you don't have to read the whole thing. Just referencing it. It's good. >> Yeah, that's what we do at the start of every meeting. Okay, cool. And everybody here is in it anyway. So, >> yeah, exactly. There we go. >> Hey. >> All right. Oh, Nico is already here. So, fantastic. >> Um, >> awake. >> Oh, wait. What? >> Don't you have a new baby? >> Oh my god. Is that Is that >> No, no, that that's that's in a few months. That's in a few months. >> Okay. Okay. You're you're well rested then. No problem. >> No. >> Yeah. I'm not yet completely uh completely flabbergasted. You you'll be able to tell. You'll be able to >> I'm quite certain of it. Yeah. Exactly. All good. Excellent. All right. Well, then let's uh dive into the agenda. Um uh who's got that first one? Suspend. Oh, suspend dormant meetings from the calendar. This was a carryover from the last meeting as I remember. Uh, and there are two meetings that are not active. And so what's the proposal? I think uh uh Neil, were you the one that that recommended we do something about this? >> Yeah. Well, well, let's put it this way. I was in the steering committee over at DIFF earlier today and a similar topic came up around a number of topics. I think the two in trust of IP um are the data modeling and governance architecture task force. Um I see John's on the call. Um so so you know we we haven't had a meeting in some time but you know it's useful to keep that slot open. Um same thing for the data modeling stuff. I will have a major update this fall. Uh been heads down in data schema work over at diff for travel. Um so major update coming on that and some more sessions in the fall but right now it's uh it's pretty quiet and part of it is to do a test with that particular project as to what the interest level is in data issues related to trust of IP. So what is our is our proposal to take them off the calendar but just not put anything else in those time slots or what what's a proposal? Uh I would I would say and I see John chiming in is is let's leave them where they are and uh you know post in the various spots like Discord etc that uh you know at this point there are no planned meetings um but stay tuned and leave those also open in Discord so that uh uh that information ga posted by the uh group leaders um uh both in the uh main discord channel for trust IP and and the particular uh topic groups. >> So should we maybe ask that they re be renamed and called holding for so that or something because otherwise people think >> that it's a live meeting and then they go to the meeting and there's nothing there. >> No, that's fair especially for people who are new who >> are not uh so so yeah I I I certainly don't see a problem with that. >> Okay. So can and John Phillips. Okay. So, can somebody send an email to support at trust rover IP and ask that those two meeting slots be named uh uh um I don't know come up with a name that implies that it's uh not active but we'll reactivate and then when it's reactivated then we can send a message to support and ask it be name changed again. David, >> sorry I wasn't meaning to interrupt. I was going to just add something when you were finished, Carly. >> Yeah, I'm done. >> Okay. But I think this is great. I think you're right. I mean, you don't want to send people to something that's not going to be uh uh active, right? You you don't want people to go down to a dead end somewhere. So, I think putting that note andor removing them from the calendar is a great deal, a great idea. I was just going to add the same goes for any Discord channels. If you see any Discord channels where if somebody showed up and asked a question, they weren't going to get a response, we should be, you know, considering archiving or pruning those as well. So, you know, I think looking at the calendar, looking at Discord, looking at any of those places where we're sending people and just pruning, you know, every so often on some sort of quarterly or halfyearly basis, I think is, you know, a good practice. >> Okay. >> Yeah, agreed. Um, I've noted an action there. Um I'm I'm happy to to do it, you know, after this meeting. Just send send them something for that. Um >> Okay, >> I'll do that. Yeah. >> Yeah, it's a support at Linux Foundation or decentralized. Yeah. Anyway, >> have speed at this point. >> Yeah, just just post that just post that in the chat if you would please. >> And the chat. Okay. And Discord. Okay. All right, good. Uh, okay, Carly, the next number four is um Yeah, >> this question. Go ahead. >> Yeah. Yeah. So, I'm in I'm in the I attend the Kerry meetings and uh and they were asking about if you're fixing a typo, non non-normative, not changing anything, you're just doing a patch, do we need to do an official 1.0.1 0.1 release or can they just change it and not mention it? >> Um the >> the strict rule is it is that any change um even a typo if it's just a a patch it's okay for it to be a patch. It's just a point one at the end. But anything you change, you should >> um cuz you know it's just >> y >> it's best practice. >> Yeah, John. >> But they don't have to change anything else. >> Uh go ahead. I >> I was Thanks. I was uh reading your notes prior to the meeting, uh Kylie, and I and I for a while I thought, well, yes, you could ignore it, but what just said is right also because most of these standards will now be read by machine. So if you made a change, it knows there's a change. it doesn't necessarily know how important the change is. So just a human might just see as a typo change but the machine sees it as something different. So uh so I think yes we do have to go through the minor release change >> and no need for any specification approval. >> I would agree if it's if it's a patch release >> it's a patch release the working groups can do it themselves. Okay. And uh and then just as a note for DOIs, so these these are now being uh put into Zenoto. So uh Zenoto only has a single version number and it just auto increments every time you make a change. So it's out of it's out of sync with whatever. And that's okay. We can call the Zenotto release the major minor. So it's 1.0. And then there's a whole bunch of little version changes that happen and those are just patches. So when they patch the next one, it'll go into the same Zenoto repository as the next version in that repository, major, minor, and then all the patches. When there is a significant change and they're going to do a major or a minor release, that would be a new Zenoto rep uh new Zenotto um uh repository and it would be called 1.1 or 2.0. And and then every update within that every patch just happens in that same you just update the document and it does an auto increment and you the authoritative numbering is what you write on the spec not whatever Zenoto thinks it is. >> So that's my Yeah. Yeah. >> Should we include a reference to Zenoto's uh documentation with regard to that? Would that be useful or is that not necessary? We we can they they so so I'm the liar. Nobody really the when I talk to people about this kind of thing it's all kind of and when I read it's all kind of messy. So nobody there hasn't been established norms so well for this. So maybe what I will do is I will write it into so put that down as a to-do. I will write it in. I don't even think that Zenotto thinks like that because their thing is like a lot of it is is like we're done. We've published the paper, we put the copy in, we're finished and it's never versioned or touched again because I'm on to bigger and better things. So that's what the infrastructure is built for. So I I think that this patching system is the I've been thinking about it for a while and I think it's the best answer for this. So that's my proposal and I will write it up in confluence. >> In Confluence or are you going to add it to documentation of of of the >> in confluence? We have the documentation about working >> to the Okay, got it. >> Yeah. So I I I mean I guess I I ask for approval from the steering committee that this is our official recommendation that that major minors are significant enough that you would get a new Zenotto folder to work in and uh and then every patch the working group can do by themselves and then they it just whatever uh uh um adding the a new patch version to a major minor release and then and then a lose a notto when you go up a major or minor. So is that uh uh is there any disscent from the committee that this is a good approach for us to adopt? Okay. And I will report back to the committee if uh there's any uh if it turns into a cluster. Okay. >> Good. Okay. >> And then I'll take the next one too because this is my discussion with Thomas. So Thomas can come last time. Thomas did an excellent job of preparing for the all member meeting. He went we went through our notes. So for reference if you'd like our last all member slide deck is there. The new slide deck is in the minutes. That's the templates to be filled out. Uh we have a list of working group chairs. I ask the committee now to have a look at the members with me and is there somebody who should be added or subtracted because then we can send it to everybody and say uh please add names or whatnot. So Drummond, Thomas, Martina, Nikki. >> Well, there's two people not not listed. Uh, they're on the other page. Yeah. >> Oh, so we're supposed to look at that page about Oh, I see. We're just supposed to check that list. >> Yeah. Steve Steve Milstein is on the second page. John Jordan, Eric Jury, Carly Haidimo, Wenjing Twice, Neil Michelle, Jonta, Steve Macau, Daniel Beheimimer, John Phillips, Samuel Smith, Kevin Griffin, Philair Fair, Scott, Perry, Mackey, Daryl, Nikki, Martinez, Thomas, and Drummond. Is there somebody else? Jan, if you give me an email address, I can add them. They will love me. >> Is Is Nikki on there? I'm not looking to list. Yeah, Nikki Nikki Hickman is on here. >> Okay. >> Yeah, >> there's there's one task force chair that's newer, Daniel Glint. >> G L I N Z. >> Yeah. >> Shoot. Okay. So, I'm going to need I'm going to need I'm going to need those email address. If if if you put the >> we're going to close the Zoom meeting and I'm gonna lose a chat. So if you can put it in the in the agenda then I will uh so >> got it. I have to >> I'm clicking over the document but it's asking me to log in to LFD centralized trust. I'm like oh that's interesting. Haven't had to do that in a while. All right. Anyway, I will I will add those there. Let's see if uh yeah SSL works. Okay. Okay. I I give up for that for now. Um Thomas, so we have the Zoom link for the call. Thomas said that last time we ran 90 minutes instead of 60 minutes. Do we want to extend that time to 90? We don't have to use the 90, but we can book the 90. I suggest we book the 90 since last time we went over. >> Yes. Yeah, I think we should do that. Um, Carly, I did do the login thing and now I'm at groups.io. Um, am I supposed to add folks try and add folks there or how does that work? >> I I think so. At one point I was able to add I'm see I'm an admin in the working chair group and somehow I should be able to add people. There we go. I've got the power of direct ad. So I I can do that if with an email and then and then they're going to get a surprise email. >> So the the groups I don't even see one for working group chairs. So I must not have that permission. >> Yeah, I had to log in. >> Well, no I am logged in. >> Yeah. Yeah. Yeah. So that so so write to support at Linux Foundation and you can ask it as an admin. >> I got it. I got it. Not a problem. Not a problem. It's all good. Okay. >> Okay. But I in the meantime then if we have Daniel Glenn's email that would be great and then I could >> I can get that I can get that for you. I put it in put in signal. How's that? >> Okay. So So here's a decision. Uh and I will communicate that. >> Yeah. So Carly Carly. >> Yep. >> Yeah. So we we'll I'm fine with the booking the 90 minutes, but we'll time box it just like we did before. >> We'll figure out what's the agenda, how much we allotting individual speakers, and then we'll figure out how much time we we really need and communicate that as we go forward. >> Yeah. Yeah. when it comes time to the because then onto the second part about uh uh the promotion uh um I'd like the last meeting then are we okay that we open the meeting to the public promote it on the links foundation event page and social media and the newsletters and uh and then when we contact the working group Thomas is contacting the working group leads and will remind them that this is a public uh this is a public invitational members meeting. Are we okay with that decision of having it open the same way we had the last one? >> Uh to my knowledge we've always had our >> numbers meetings open to the public. >> Yeah. But it was it was like we talked about it versus we advertised it. >> Okay. Oh, I see. Okay. Well, >> so we will advertise it more broadly. >> Yeah. >> Okay. Last time we live streamed it on YouTube and we had some people there. >> Are we okay with live streaming it again? Okay. >> Yes. I I'm going to shout to the world. >> Okay. Um Tomas will host the MC unless somebody else wants to. I asked Tomas. Does somebody else won't really want to know? >> Oh, no. Are you kidding? He's the best. >> Yeah. Okay. Okay. So, Thomas will host the meeting. He did great job. Structure the slides from the last call. Uh uh and then um time at the end to discuss IIW and stuff like that. And um uh let me put in the chat for everyone then um if if anyone has any suggestions to what they would like to change on the structure. So Thomas will send out a email to the chairs and that's why we're updating the chairs list. Here's the other thing that we are doing and that is we asked to um harmonize our reporting at the all members meeting with our reporting to the technical advisory committee. So this means that in the email that Thomas is asking for for in to people to contribute to all members meeting is also to get information that we would report to the TAC. So that then I was uh so we talked about that with Thomas and then we talked about that in the reporting meeting. It was a 17-minute meeting this morning. Uh um and we came up with so this is the we can share this with Tomas when Tomas emails everyone and if people would like to use this prompt to get evidence for the review or the human if they just want to type it and then we will ask everybody to contribute this information since they're getting ready for the all members meeting and then to go to into the TAC So this is my request to the committee if we um if we agree with this. I mean it had been Drummond and I who had been doing the TAC reports and it was a bit of ah a scramble. So >> right, >> if we can piggy back all members in TAC, then we might have more less scramble and more accuracy or more representation. Okay. >> So, >> and and folks who want to let their agents do the work can do it. >> And if you want to let your agent do the work, yep. David, we haven't heard back from AFAQ. You have not. So what was the status? I remember. So I connected you to Bobby and Jean Luca over email. Did you end up Did they set up a time to >> No, they said they were going to discuss and get back. >> They sent me Let me see if I can find a link. Maybe they didn't communicate it clearly. They sent me a proposed document. >> Oh, okay. Has anyone here set up an LM, a local LLM with a rag? a local one. No, I haven't. >> Here, I'm going to drop a link in the Zoom chat, Carly. Um, this is what Bobby sent back to me, but yeah, I'm I had Sorry. I I had just assumed maybe she had reached out to you or Jean Luca reached out to you to set up a call to review this document. >> Okay. No, >> I can I can ping her again or you could write her back and say, "Hey, David shared this dog." I mean, they're it seems like they're offering to help. >> Yeah. Yeah, that would be that would be okay. I will I will follow that up then in we have another meeting with the um reporting Nicholas. Uh I just think in in terms of um you know aligning efforts across the different working groups and and projects here um something that could be useful at least as an input for a rendering of the rag would be the specup t uh um u glossery right >> what do you think about that >> yeah well we were uh in the proposal I I don't want to take too much time. But in the proposal, we were like what we would want is to identify the sources of information that are all of Linux Foundation decentralized trust is storing where like GitHub repos uh specup t documents conferenced here. I thought this was more internal understand. >> Oh yeah. Yeah. Okay. So if it if I didn't have it in there then yeah the specupt specs for sure and and and the GitHub where where where they're hosted and stuff like that. Awesome. Okay. >> I like a lot. >> Thank you. >> Okay. It it'll be fun. It'll be interesting to set it up. Okay. Uh Daniela asked for a slide on how to join LFTT as a member and uh um I'm assuming that I will tell Thomas to put the slide in unless I have a reason not to from the committee. >> Oh yeah, there's no reason not to. Yeah, that makes sense. I >> thought Okay, sure. So, >> uh, Drummond, do you want to, now that I've covered my Tomas meeting stuff, do you want to take it over? >> Uh, um, >> number six. >> Yeah, number six. Okay, hang on just a second. I got to get back to it. I was trying to find, uh, um, second. >> No problem. >> I need to get back to that. I got to get back to the right one there. Uh oh. Well, the the so this is just um we were number six is basically we were talking about what are topics that we anticipate folks if we're talking about highlights. I'm actually put this in here. I'm going to say potential highlight. Oops. What just happened there? Uh oh, my cursor was in the wrong place. uh possible highlights and this was the list that we had uh um we're just sort of ticking through and going well there's a big one about um >> GSP >> approaching yeah final we now integrated rev 3 so we're on flight path to a 1.0 Um, so Wjing will give an update on that. I'll I'll give it update on DTG. Um, Scott will cover COG. Uh, we figured Carly always have great news to support on the latest on by-size trust. Uh, Daryl's got a lot um after GDC to report on trust registries. And then we have a brand new working group uh virtual private knowledge, right? Um, assuming that Rea as long as we he knows that we would like him to give an update on that and that's I mean that's a lot already >> and then and then and then whatever Carrie what does Carrie I know they're working on some >> like they did webs they're still working on that >> now they've got dossier thing. >> Yeah, the dossas are something that um are will be pivotal um and we might even have something uh an interesting use case for for dossas in particular beyond just um um telome >> who uh Nico who besides yourself who else would report on that or it would just be you >> I think I think Daniel would be interested in doing that that's why I mentioned him above he's uh the as far as I understand if he's not the chair of the dossier revival dossier working group I have no idea who is um >> they always stick it at the front of uh uh at the front of did web and that and then uh Kent I don't know if Kent can also talk. >> Yeah. Oh yeah, >> Kent. >> Yeah. >> Is Kent a chair? He's not here. >> Let's see. That's a good That's a good find. Um >> he does a lot of work, so I might as well stick him on the list. >> De facto. He's a chair. Yes. >> Okay. So, if somebody can find his email and tell me and then I will add him. >> I'm not sure I have his email. Hopefully, um, Nico, you do or someone does. >> I've got it. >> Okay. >> And, uh, oh, the TAC reporting I already covered. Oh, sorry. Um, back to all members meeting. Uh, any other thing that I'm that we Well, I think we we created this in advance before you would talk to uh Tomas and and we're going to go through this process to get the updates from everyone and I'm just we were >> so I should know what we're looking for so we can >> Yeah, exactly. And that folks can, you know, be preparing for it. Um >> Yeah, >> that's good. >> Mackie. Yeah, actually I just wanted to add like maybe as part of the um you know meeting maybe we can shed light in terms of the latest uh conferences attended by our members including GDC for example. Uh that will be good um at least you know like key takeaways or something of that sort. So, so Drummond to that note. >> Yeah. >> The in the last all members meeting we started we started with uh market updates now >> or would we end with market updates because to make sure that we have time to cover our working groups? Uh yeah, I think well as long as we make sure we have enough time um I I do think it's good to start with overall market updates rather than the end. So >> it's like >> yeah. >> Okay. Uh so uh Drummond me and Scott not that I have much of a market update. All I have is PIDFest. That's my uh that's my market update >> actually >> because I stay at home. >> That's a blessing. >> Well, I would include some market updates as it relates to COG within my fish. >> Yeah. Yeah. So, so we the only ones we put up front would be, you know, that that are not working group specific. So, you know, general stuffs. Um, then we'd save that for each working group or or task force that's going to report on it specific things. Yeah. Okay. Yeah. Okay. I I can help prepare that. >> Awesome. Uh, the TAC reporting. So, I talked about that as part of the all members meeting data prep. Um, and then the update on the A report AI reporting. So, we also covered that just now. So now I have a uh a proposal from AIFAQ and that is being discussed if you would like to join that meeting that's there's a reporting task force that meets on Wednesdays uh and we have a discord channel. So um and that would be that would be the pipelines the rag and and implementation here at LFDT. Um so therefore we have covered our uh bulk of the agenda. >> Um yeah we have covered quite a bit. >> Can I say in record time I think this is the fastest meeting we've had in all my time in the uh the steering committee. >> We're we're not done. I I do want to hear about this prohuman assembly but the uh the meat. Is there uh uh um is there any other meat here? >> Yeah. >> Before we get into that, >> I I would um so so it's it's not quite ready uh yet. Um but by this steering committee next month, um Daniel Hardman and I may have a proposal for a new working group. Um >> the dossier one. It's built on dossas. Yes. >> Okay. Okay. >> Yeah. >> I mean, if it makes sense just to pull it into dossas, then you know, so be it, right? Um, >> a work a working group or a task force. >> I don't know actually. Uh that's what I need to >> if it's under carry because a working group needs a a working group needs steering committee approval but task forces can be spun up within working groups with just working group approval. >> Okay, I see. So in this case you don't you don't need steering committee approval to get a task force up and running. >> No. And the other reason that it's harder to get working groups up is that there's a fair amount of legal documentation then that we have to go through as well. >> Fair enough. Okay. Well, then we we'll see. I mean, h one question and I'll leave it be. Um how is the process of um does a task force I mean I guess this did happen right before a task force can become a working group if if the work warrants it. Yes. >> Yeah. >> Okay. Yeah. Yeah. Yeah. It can expand. >> Excuse me for I know this is all redundant, but uh it's good to refresh on all the details. So, thank you. >> Yeah. Because I I did not participate in the new uh working group uh setup, but I did see there was a bunch of meetings and signatures and stuff like that. That's for the VPK >> working group. So there it wasn't it wasn't there was a a significant amount of friction in there. Okay. >> Sounds like fun. >> Okay. Um and then Yep. Okay. So then and and then Oh, what due date should we set for TAC info and TAC info and um all members meeting? content. >> You mean you mean for things to be submitted? >> Yeah. >> So, where are we now? >> Should I should I set that date with Thomas because he was the one who um >> Yeah. >> I'm imagining it's going to be a week before and I will find out from Thomas what fits with him. This is that's Wednesday the 28th and a week before is steering committee meeting. >> Uh yeah. Yeah. Go ahead and set up it with Thomas. >> Okay. >> I will I will connect with Tomas and set that up. >> Cool. >> Okay. I I want to hear about this prohuman assembly and if anyone else has uh things that they've been doing. >> Robin, do you have the assembly update? >> Tell me about your personal conversation with Bernie with Bernie Sanders. >> All right. I've joined my I took a bunch of pictures with my iPad. So, I've joined the meeting on my iPad and I'm going to switch back to uh Zoom and see if I can share my screen here so I can show some of these pictures. Um, the easiest way to let me try this I want to do photos. There we go. And I can I show all the photos. I'm trying to I've never done this on the iPad before. Try and uh can do that. I'm going to show. Let's see. Oh, what? Select photos. I can't just share my screen. Oh, wait a minute. Let me try this again. >> Neil, share screen. >> There you go. Or broadcast. Oh, that's fine. Okay. Interesting. On the iPad, it considers it a broadcast. Is it Are you seeing anything? >> I'm seeing a screen broadcast and it's a a thinking the the the Apple thinking uh twirly thing. >> Not doing anything. Yeah. Listening. All right. I'm stopping that not working. Screen broadcast two one zero. Oh no. That's weird. Same thing. >> Now I'm just I'm just trying to get it to go away. >> You're sharing my screen. Great. >> Oh, now I'm sharing my screen. Okay. What are you seeing now? >> I I see uh you started your screen sharing, but nothing is showing up there. >> Are you now seeing bunch of photos? >> Nope. >> Nothing. It just says that you started your screen sharing. But >> actually, what am I doing? I've got my I've got the uh meeting open here and it's not that's not this is not working from showing this stuff. That's funny. Okay. Well, maybe that's not going to work. Um yeah, I really did want to show you some of these. Uh now, let's see how I can turn this off. >> Maybe you have to email them to yourself. >> Whoa. Stop share. Yeah. All right. Watch for that. Is it still screen sharing? >> No, now it's uh off of screen sharing. >> Okay. All right. That's not going to work. Um I do have a few pictures on that. I took >> So, what what is the prohuman assembly? Um, I got to turn this second zoom off. It's echoing in the background. >> You got the echo? Yeah. >> Yeah. Good. Um, it was only it was only called it or they came up with the idea about two months ago. Was put on by something called the future of life institute. Um, which is I think futureoflife dot or I'm not sure. Anyway, you can you can Google futureoflife institute, but you couldn't um you couldn't Google anything on on the uh future of life assembly. I mean the uh prohuman assembly until uh let me see if I have now pro. Yeah, there it is. Uh prohuman assembly.org. So yeah, I want to go to that one. Okay. So, I will share my screen just so you can I'll put this in the chat and you can see the uh where is it? Over here. They did finally put up this website just before they they started it. Um and uh let me share my screen here and we can take a look at this that guy. There we go. All right. So, um it's in Washington DC. It was two-day two-day event. And um the future life assembly is the two chairs of the event are two um professors. Uh one Max um I don't know if he's listed here on this on this. Anyway, Max, um, no, he's not. This is one of the, uh, um, I can't remember his his last name right now. Um, Wegman, something like that. Anyway, he's a MIT professor, um, a physicist that also is a professor of AI. And the other one, Anthony Aguiri, is Caltech professor, um, also AI and other things. And uh two of them um again were the chairs of the event with a a third woman from I think the president of future of life institute uh woman I don't remember her name either anyway they were the the um MC's that ran the event and uh the f the first day yesterday was um a series of keynotes and panels all about um how AI can be um prohuman is is the simplest way to put it. How how to ensure that it's under um uh human control and that it's good for humans. And yeah, the the the headline of all of them that it started out with three congressmen uh I believe it was two Democrats and then one Republican. So they made sure it was bipartisan throughout the whole day. um and they each talked about the legislation um all of them were co-sponsors of legislation starting down the path of regulating AI. One of the major themes of the whole thing was um that AI is the only industry with um uh you know call it global danger potential on the same order of nuclear weapons um uh nuclear power um you know airlines uh um and and air traffic control um uh some aspect you know nuclear missiles I mean all of those things have heavy regulation around them and yet we're sailing into this world of AI and um some very very scary stuff. I think folks have seen me um share the the book called if anyone builds it everyone dies. Um that book is not a joke. There were uh there were actually some folks talked about in the book who were there and so the the uniform theme throughout the day was we need to start regulating AI even though we have a certain president right now who doesn't think that's necessary. Um the the goal of the event was to energize uh a lot of folks and and politically as a movement to uh to recognize we do need to regulate AI. And the the most powerful thing, the thing that got the New York Times uh story um you can see the speakers here, but the New York Times story is the one that I um just found. I I haven't read it yet, but it was uh not there. I want to put it. Go ahead. Um uh Nico, while I find this >> Oh, I just uh you know, it's Eliza Udekowsk's book, If Everyone Builds It If Anyone Builds It, Everyone Dies, is uh is interesting. I I want to just talk about that for a bit, but I I would rather you continue. >> Um yeah, let me let me find this. Um that book if you have read it uh was a uh a really um uh it it was certainly the theme of the event uh was very much okay that's it is a reality um the kind of danger that is discussed in that book and a lot of people go oh come on really how is AI going to uh you know lead to the extinction of the of the human race if you haven't read that book. I highly recommend it because it actually the last third of the book spells out a scenario and and you might find it implausible, but I've talked to folks in the industry and there were folks on the stage that said, "No, this is entirely plausible." Including I actually do have a picture of this one. In fact, I'm just going to switch to it right now and I'll share this. So, this morning there was a special speaker um that some of you might recognize if you're deep into these things. Uh oops, there you go. That guy right there is Jeffrey Hinton. Uh pretty widely acknowledged including at the there's a fabulous um biography of Deis Asabis, the founder of uh Google Deep Mind. Um and in that biography even he acknowledges that Jeffrey Hinton the uh right there is the father of much of it and he worked very closely with Demis and so the guy interviewing him right there is Max um the co- um the MIT professor that is co- um facilitator of the whole event and he interviewed him this morning for about half an hour and one of the questions he asked is hey people think that it's not you know that that it's a wild fantasy that AI could kill us all. What do you think, Jeffrey? And he came back in about a fiveminute answer and said, "No, this is really a serious issue." And he explained uh more deeply what happened with the ai and hugging face event in August uh of the agents breaking out and how they conspired together and agreed to cheat. and he he he used the term um these we need to start considering that we have created alien beings and uh you know this isn't a conspiracy theorist or a crackpot this is the father of most modern AI and he was very cogent as salient and saying you know it is the fact that we're that we're uh driving a car at top speed uh towards a cliff and only talking about um whether we need to um you know apply a little bit of brakes, you know, is is insanity. It's like he said that car has no steering wheel. If we don't do any regulation, we have no steering wheel on that car. It doesn't matter if you start to apply the brakes unless you're going to stop the car completely and nobody thinks that's feasible, you're still headed off that cliff. So, um, anyway, it was it was very, um, compelling. We heard a lot of stuff yesterday, but this morning when he, uh, in this interview because I guess he wasn't available yesterday, it was like, whoa, that's, you know, uh, so anyway, um, and I'm not trying to be a doomer. There was a lot of, you know, reference. No one is saying, oh, no, let's not let's kill AI altogether. Um the major theme is let's really figure out not just regulation but also how to build a prohuman AI and uh I think I have one other picture in here that I will share that yesterday was all box today was a workshop that was the particular reason I was invited and here's the picture of that I took oops come up here of the name of that workshop you seeing this is this coming across Is it still okay? Good. Yeah. Um, yes, it was Chattam House Rules, but I'm not sharing anything about the workshop other than the name the workshop. Um, and uh, so there were turns out there were six workshops today. I didn't realize I ju I only knew that there was this one which was about 30 of us um working on all aspects of what they were calling the prohuman AI tech stack. >> Um then >> sorry just to interrupt to remind people that this is recorded and put on YouTube and even if people are usually not going through the steering committee minutes the AIS are. >> Oh yeah. No, I understand. >> I was just double checking. Yeah. you're seeing you're seeing the lower part of this uh which is that it's under Chattan House rules. So I'm not going to say anything about um uh you know or attribute anything to anyone. Um I mean the fact that Jeffrey Hinton was a speaker this morning I think is is already um uh public. Um and of course the New York Times article covered um the major speakers yesterday. I mean, the headline yesterday was the fact that you had Bernie Sanders and Steve Bannon talking back to back. Um, and they neither one dissed the other at all. Um, Bernie didn't reference Steve Bannon at all, but uh, Steve Bannon did reference the fact he was following Bernie that well, they have different opinions of the current president, but they completely agree about this topic of AI regulation. Um, and that's what, you know, got the attention uh of the, you know, got the New York Times article uh about the whole thing. Um, anyway, for for for this room and for uh the recording, I can just simply say um that um I made a pretty strong case that hey, if you're going to build a prohuman AI tech stack, guess what stack you should build that on? And I actually showed people, guess what, a a classic picture, John uh Phillips would be very happy, the highle picture, the four layers of of the of our stack. And people like, oh yeah, that makes a lot of sense. So um anyway, we're in a very good position to help with that. There's going to be a lot more that happens. Um uh there is a conference called the global AI summit that's happened I think three times already but the next one's in in Geneva next June and so there's a whole work track that they want to prepare to actually um show a prohuman AI tech stack running by that time and there's certainly a bunch of stuff that we're working here on at trust RP um in multiple working groups that we could show as part of that. So, I'm wondering if there's anything else. Um, and I'll just quickly look and see if there are any other pictures I want to uh Yeah. Um, no, those are the main ones. I took those on my phone, so that's why they're here. The other ones I can't uh they're on my iPad and I can't show them right now. Um, so that those are the main things. I just wanted to say it was I think there were about 300 folks in the room yesterday and then like I said about half that number were in six different working groups today. Uh I was told in the policy group there were about 60 people. I mean it's Washington DC and this is about the need for AR regulation and so this was you know the Washington policy corps fullon full at it. Um and uh but uh but it was um the the consensus around the need for this was unbelievable. Steve, you have your hand up. Go ahead. >> Um yeah, this is an interesting Thank you for this summary. This is super interesting. Um, today I read a article where it talked about the CEO of Anthropic is calling for a pause on AI while the CEO CEO of Nvidia says to keep going forward as fast as you can. Um, and so I'm curious, since federal regulation tends to favor large companies and disadvantage small companies, did they happen to say what they're advocating so that it doesn't put small companies at risk? >> Um, I I'm happy to say that was a major theme throughout the whole day. um every discussion around policy and regulation. Um they even talked about you know regulatory capture by the large companies and so I mean no surprise the strongest voice on that very point Steve was Bernie Sanders. Um and uh he even they they referred to several bills. I think I'd heard of one or two of them. Um but that they were um you know they they really really wanted to make sure as much as possible that they were emphasizing uh open market, open models, decentralization and a level playing field for all competitors. Uh I don't want to characterize the event as anti- big AI. um you know they're they're recognizing in reality that those companies are there and that but they are anti- race to AGI they are I mean they're they're saying um Jeffrey Hinton put it I actually I wrote it down uh but I can paraphrase he said um a race to AGI is like um or or trying to win the race to AGI is exactly like trying to win a nuclear war. everybody loses. >> So that that's interesting. But I I guess part of what's I I've been like all of you. I I've been following this please slow down AI call for the last little while. So what did what did they say about um if they're successful in getting the US to pause that doesn't really affect other countries say in Eastern Asia somewhere that might not be putting the pause. So, how does slowing down one set of participants while not slowing down another make everything safer? So, um that subject was also not only not avoided, it was talked about by at least six different speakers including Jeffrey Hinton this morning. And uh they pointed out the same thing that I I was like, "Oh, never thought about it that way." Um, and this isn't so uh we don't have to beat around Bush. We're talking about China here. And several of the congressional and and and Bernie Sanders, actually all of the um political speakers brought up and said, "Yes, everyone says if we regulate and China doesn't, we're going to lose the race to them." All of them made the same point, which is, uh, think about that. Um, which of the two countries actually cares more about keeping control over its citizens and control over their technology? >> Well, the side of that is is China makes products they sell to us. >> So, if we've paused AI in the US market and China doesn't pause AI in the US market, it creates an imbalance. Yes, there was discussion about that and well, guess what? Regulation is one of the things that could deal with that. I mean, I'm not a big fan of uh, you know, trying to cut down global trade on that thing, but again, that was compared to, well, what do we do about nuclear weapons, right? Um yeah, other countries can create those nuclear weapons and but controlling the flow of those nuclear weapons or or the movement of those nuclear weapons is something we have international you know trade agreements about and inter international disagreements about. So part of what and it it came up a number of times and it is it continues ever since I was go at GOIM um in Paris um in the spring which is largely u Chinese uh sponsored open source um conference. China right now is out competing the US and the rest of the world in open source and in particular open AI models. And so we also need to figure out about that because if they continue to make those the most attractive economically, you're right, it's they're going to attract a lot of folks to use um uh you know Chinese developed uh AI technology and uh I mean they can do that regardless. So it is another open question. Uh I guess Nico go ahead. you know on on that uh sort of bend um right the Chinese open models if you actually look at the um if you if you ask them in English uh what uh what model are they they will often times call themselves claude right uh so so the creation of these uh these sort of open source models are in part because they the Chinese are the willing participants who are distilling uh closed source models like claude and potentially now chatgb BT and Astra. So it's almost like you know the the intellectual blowback right from from keeping models close source in and of itself which is the problem. Um so in that same vein I think a lot of the existential risk of of what we're seeing with AI is the uneven distribution of it. Right? the technocrats that currently hold the keys toward um and and you know you see this as well like the these people that call themselves effective altruists and anthropic right think they know better than anyone else with with regard to not just artificial intelligence but in general now what what they one could claim intelligence. So I think the the risk comes from the fact that um artificial intelligence as we know it does not resemble the actual distribution of real intelligence. Right? Intelligence itself is not something uh anyone can hold a monopoly on. Um and so in this case the systems that we have and the models in particular are a monoculture right that tend to go toward centralized administration of this uh of these machines which we can infer on. Um so and and you're you're seeing actually a lot of interesting developments in terms of and you will see even further the the massive multiplicity and and and and diversity of what and how you can actually uh create artificial intelligence. There's one that just came up uh recently my radar two days ago called type safe uh which is a completely different model uh which which has radically different economic assumptions right uh be and and and reduces the hallucin hallucination rate down to virtually nothing right a homespun American model right and will probably be um generating its own class of open source invariants so so so I think a lot of this discussion has to be tempered with the fact that we are seeing a a a sort of feedback loop about everyone using the same type of model which has its own weaknesses and that as the ecosystem develops you know we just really have to just survive until the point that we can uh have a more nuanced understanding of the governance of these things um and and and really like you're saying Dringing bringing the sort of infrastructure at scale um that can actually administer a humanfriendly uh cybernetic interface layer with not just not just LLMs but a whole host of of new types of artificial intelligence we have yet to see. So I don't know I just that's the way I see it. I'm looking at what Steve just put in the chat Bill Joyce quote the standard definition of AI is that which we do not understand. Yeah, that point was made a number of times which is that um and and Jeffrey Hinton said the same thing this morning that the models the frontier models that have been created now are beyond the understanding already of the of the developers. Um which is why you know when people talk about okay it can get scary as like yeah. Oh one more point that I thought was um I think I I ended up posting like eight posts to LinkedIn. I was treating since I don't use X anymore. I was and and I didn't think Blue Sky was the right audience. I was posting this all linked down. Uh but >> post it on Blue Sky. I want to have stuff to read. >> Yeah. No, I should I should >> I'll like your post, Drummond. I'll like on Blue Sky. >> I should >> I should be putting more on Blue Sky, but um it was uh pointed out that uh the fellow um and he's not the only one. Several more have left u in a resign from open AI or anthropic either one weeks or months ahead of IPOs that would have made these these folks millionaires or multi-millionaires right so when the top some of the top AI safety folks at those firms are leaving and publicly you know um ex explaining why and you realize that they have just given up millions in stock options. Um, you know, you can tell that's like they're they're serious, right? Um, yeah, I thought that was a really good point. Anyway, I'm not I'm not here. I'm just trying to report the the most important thing from a trusty peace standpoint is there is universal agreement. Well, we need we need AI um infrastructure to be trusted and every the point about well you have to have proof of personhood for that is just like table stakes now. Everyone gets that. So um so from you know the standpoint we didn't I was I think the only one uh directly involved uh with with with the trust that was there. Um but anyway, in the in the techstack group, it just seemed to be um assumed, okay, that's a really good infrastructure. We should be from a trust standpoint building on that. Go ahead, Carly. >> Um did you see the uh 404 article a day or today or yesterday, something like that? So they the guy reported about how now all these AI agents are soliciting him endlessly with pitches and jobs and stuff like that. It's like before they had to type it by hand, but now you have AI agent supported trolls and spam and everything else. It's just uh when I looked at his inbox, I decided that I thought it might just be better to turn off the computer if agents are allowed to send emails. Yeah, this question of provenence now for any form of communication >> is huge. >> It it almost needs proof of proof of work before you can send me something before you had to type it out and think and now you just have to spend a token or two. >> Right. But when I talk about the prominence, if I I where I think we're headed is um is is is filtering based on trust relationships, right? >> Yeah. >> You just if you don't have a trust relationship there, you're just not going to get anything. And and so >> I mean >> I will make this point because I'm like >> I'm I'm sitting in the room, you know. Yeah, exactly. I'm sitting in the room. I'm going >> whoa. The fact that we have been at this for seven years building this stock and building this infrastructure and now suddenly it's needed everywhere. It's like wow we were doing a pretty good job seven years ago figuring out that we needed this. Yeah. Um and as >> Exactly. Because if we were starting now, we'd just be in the noise with so many other folks saying, "Oh, we have the answer. We have the answer." Right. There are at least a dozen proposals for aentic identity out there and uh to my knowledge we're the only ones sitting here saying well not only have we built a decentralized trust stack from the bottom up but we're building solutions for human device and agentic identity and community and networks and it it's all one rational infrastructure and um and you know again I'm biased but man we just we just need to move this market as fast as we can. Yeah, lots of hearts. I I >> remind it reminds me of a lot of the conversations I I was involved at GDC around the open verifiable communications and um the need for for appropriate evidence. Yeah. Yeah. Neil just put in um the point the challenge AI agent successfully is spoofing trust relationships. I can't tell you as I describe to folks how um you know how proof of personhood is going to work um in in a decentralized trust graph how many folks lock in on that and go how do you know that you're actually you know that that people are not going to be um you know that that you're folks are not going to try and game the system. I'm like, folks are absolutely going to try and gain the system, but you need to build a system where the incentives are for the communities to protect against that and for people to protect. In other words, if everyone's got the incentive uh to protect against that, then the system's going to be self-healing instead of um the current situation we have. Now I'm I'm going to channel my good friend Scott Perry here and John Phillips to say not just 80 95% of that is governance. Oh look here's uh Mr. Phillips. Uh you know John you can't actually speak on these calls. >> I I can now. I I've been walking the dog and doing all sorts of weird things, so I I couldn't sort of be on online, but um yeah, I think it's a hard question to ask ourselves from the governance side of the stack. Uh how does our governance thinking stand up against the current era of AI? Um and I kind of wrote a little bit in the chat there along the line of thinking that we I think well obviously mostly uh Scott but wrote very well about the idea of governance of issuance of credentials, verifiable credentials in all sorts of forms. that's essential, but is it enough? And I think I think it's not enough. But I I worry about going too far up the idea of the concept of trustworthiness because one person's trustworthiness is another person's riskiness. Um, so there's a sort of there's a there's a there I think there's an emerging kind of issue or opportunity to think about governance in in the broader context of the era of AI. Um, which is exactly what you've been listening to for the last two days by some very worthy speakers, Drummond. 100%. I mean, I was so happy in that meeting to basically go, "Oh, wow. This is a governance meeting, right? It's two days prohuman uh AI. It's like we didn't have to preach to anyone about the tech. If anything, the tech was under reppresented and under discussed." I was a little bit like, whoa. They're just sort of assuming that, you know, we're going to have the answers for that. But boy, did folks, it was all about governance, John. It really was. Um now whether they know how we're going to apply that and what to do about you know a lot of the things that we've been working on here that's that's a good question but at least that group really get really got that. >> Yeah John >> I I was just going to say we the first time we've talked since GDC trust registries were one of I mean Nico others were there they were a thing this time right? They definitely are a thing. Yeah. Um, yeah, it's been a rather engrossing thing for me for the last 18 months. But, uh, uh, yeah, I think Nico's point in the chat's also really important and I guess it's a a centering idea that, um, the evidence is what you decide upon whether to trust something or not, right? So, so, uh, then that raises questions about sufficiency. Uh, what's what is a sufficient amount of evidence? Um, and so on. But I think that's a better path than sort of some sort of self-declarating algorithm saying I am trustworthy, trust me, >> right? >> that that's doesn't work, >> right? It's it's kind of the um the the irony of having self- sovereign infrastructure is that you can't declare yourself self- sovereign. You have to kind of it's a collaborative emergent thing that emerges from, you know, what what you're talking about, which is this sort of nuanced decentralized set of people coming to their own conclusions. Well, I mean, you can declare yourself whatever you'd like. The question is what do other people think about it? >> That's how we value it. So, that's yeah, that self- sovereign always came up against that. >> It did the term, believe me, I've always been torn about the term um you know that the upside is it gets the fact that you are in a control position, right? Um from an individual or I like to stress a community or what whatever the the Yeah, >> whatever it is. The downside is it makes it sound like you are self-declaring, you know, we're all in agreement about that and that's why I think the term has only gone as far as it as it is. Um, I am I I do think it's interesting though. I have not heard the term the specific term prohuman um used in the in the context of, you know, the whole AI question before. And when we first came up, oh, we're going to prohuman assembly. I'm like, you know, uh, is this I didn't even know what was it was, you know, technology related. Um, but after two days of listening to it and understanding why they chose that positioning, I'm like, well, that's that's pretty and and by the way, that's a political positioning. Um the timing of this event is very specifically to galvanize public opinion in advance of the US midterms. >> Um yeah, there's a they had some some interesting statistics and evidence about the backlash uh largely on data centers. I know we've all been reading about that, but I was like, whoa, this is really starting. I think it's going to be a major factor in the midterms. I've seen it firsthand. Um, one of the lakes in Texas had essentially no um, building codes and they were going to be hoovering up, I think, 1% of this specific lakes's water per day in order to cool the data centers without really getting anyone else's approval. And it was a huge hot button issue. the governor actually had to come in and and totally revoke the state standing of their ability to build this data center which was uh yeah so it's for sure happening. Oh Dr. also I want to say I think the interesting notion of having a prohuman assembly is it kind of feeds in the other thing that you mentioned which is that um we might be building things that are like kind of alien right because to have something prohuman you also have something that's not human. Yeah, I was when uh Jeffrey Hinton multiple times used the term, he didn't call it an agent, right? Right. >> He called it an alien being. >> And you know, a lot of folks have said, oh, let's not anthropomorphize, right? His point was, oh no, let's make sure it's the opposite of that. It's an alien being, right? We don't know how it thinks. We don't know how to interact with it and we should treat it as you know more like a you know shall I say u um certain alien movies. >> Yeah >> that's that's >> I know. Yeah it's just somebody you know going out and doing work but this is not >> yeah just just doing stuff for it. Right. The other other theme that was just overwhelming was um you know folks have not been taking the job loss thing seriously enough. Um yeah >> there's go ahead John you about say something. Yeah, I I guess obviously I'm sitting in Australia right now and just as a is it a reassurance? Maybe this is a global phenomena, right? So So I know it's an intensely felt experience in the US for obvious reasons. It's also here in Australia. We've got our own data center initiatives that are being pushed by the federal government u uh which has tensions with state uh governments here in Australia around the idea of whether they need renewable energy or not renewable energy, whether they can rely on sort of uh gas fire power and so on. This the same sort of environmental tensions. We've just had yesterday an issue where secret documents were being created around the weakening of copyright laws in Australia to make ourselves more friendly to AI kind of trollling of information. It's the same the same sort of dilemas. But I I I I posted in the chat a little earlier an economist article on uh on an opportunity and I I want to say this in an apolitical way because I know this is a very political kind of conversation often. Uh there's a meeting between America and China, the two leaders, fairly soon, and there's an opportunity, were they to seize it, to agree a an approach to regulation of AI that might be worthy of a Nobel Peace Prize. Uh I just put that out there. >> Uh yeah, I know, I know, I know that reaction says a lot, Dar, but think about it for a second. Uh maybe it would be right. And and that >> you know what, you just might have something there. uh >> president >> it's it's actually a professor from uh from uh New York University who who put forward the idea it's in the article I linked but but the other thing about jobs um the economists also ran a ruler across all the job market in the US in particular and their net contra their net consideration right now is there is a gain of 100 million jobs um not 100 million sorry a million jobs so far in the market they it they they you know they they don't they are political kind of publication. Um, they're thinking so far as it may not be the the death of work that has been forecast. Um, and uh, and there are other ways to think about it, but they're not they're very much not a pro- AAI protagonist particularly. They're just a bunch of economists, the dismal science, >> right? John, have you heard of paradox? >> Yes. >> Okay. Okay. Cool. Yes. >> Yeah. It keeps I keep hearing both sides of that too because the economic case for um you know shareholder pressure on short-term profits is like oh if we can cut our costs dramatically. Um but then of course there's all the problems that creates. I think it's still it's still um up in the air. >> Too early to tell. >> Yeah. One of the one of my favorite um again somewhat political phrases from years ago now was um a prior prior leader of China was asked what he thought of the French Revolution and his reply after some consideration was >> too early to tell. >> And I think it's too early to tell for AI at the moment. >> Yes. Yeah. The one nice thing is what's unanimous is oh well we it's got to be trustworthy, right? If if if we I mean it it it almost felt like wait a minute the thing we're all about building is already being taken for for well not I don't want to say completely for granted but sort of like oh yeah okay we have to have that and I'm like I was thinking the background of this um workshop today is like I don't think you understand just how much that is involved there you know that how big of an assumption it is that yes okay we can just deliver for this infrastructure because it's a big deal. I mean, we all know that. But anyway, um >> I I'm attending a conference on pits where it's been like decades and slowly this tiny piece of infrastructure is chugging along. I don't know how it's going to get implemented for a for a much bigger problem. I mean, we've been working at this for seven years and it's just starting. Yes. But but seven years ago or no time up until I'd say the last year have we had the kind of demand produced for. >> Yes. >> Yeah. That's that's what I keep referring to as the much bigger wave we're now surfing. >> Yeah. Yeah. Yeah. Absolutely. And I the I think the good news at least, you know, coming fresh out of this workshop today is um and and the nine-month plan they're talking about to to set goals for what they want to announce at uh the global AI summit. Um they really want to use the pieces that we're bringing forward. So, I think, you know, from the standpoint putting on my first person cooperative hat, we're seeing uh a lot better routes to funding. >> Where are the big players not? Why aren't they showing up here? like if there's trust and stuff like that, why isn't Anthropic sending somebody to our meetings or or or Google or you know like >> do they I I'm assuming it's something like >> okay >> they don't need to uh do that yet >> they don't they need to do it yet so they're not going to >> I have a much simpler answer for you Carly. >> Yeah. If you're a very large platform of any kind. >> Yeah. >> Where where is the center of trust in your platform? >> You >> you why would you want to encourage anyone else? >> You don't. You don't. That's true. You don't. It's it's against your interest to build to make anything other than you must control it all. I mean, that was the metaverse play, was it not? You were going to come into our ecosystem and we were going to run it all for you and pat you on the head. >> Which is why I think, you know, we're seven years into it. Is Google, Microsoft, Amazon, Open AI. Uh, >> no, you're right. you know, are any of them are any of them even members of LF decentralized trust, >> let alone trust over IP? >> No. >> Yeah, they're over there working on the standards elsewhere in the Linux Foundation, but they're not they're not fans of decentralized trust, and I don't think they >> will be until they have to be. >> I think you're obviously you're you're absolutely right, Drummond, in in the current current context. I think in in individuals there are people who are very much focused on this but they don't have the sway over the whole platform and I give evidence of the fact that diff was founded partly or almost mo mostly by Microsoft >> back in the day Microsoft had a team of people providing the entra platform as it was which was once a verifiable credential sort of focused um solution um and that was a team that was eventually dismantled and sort of sucked into a marketing blur the round mantra. But yeah, they they do have people that are good. >> I I think one ironically if there's someone else that's a giant root of trust >> who is or not all of them but many of them will be in favor of decentralized trust and that's governments that actually believe in democracy, right? because decentralized trust does not threaten them. It actually reinforces democracy. That's a that's a constant thing that comes up. Yeah. Other governments that are not democratic um may not be as interested. Um but anyway, I I I certainly left this conference thinking that um their concerns about AI specifically um may open more pur strings in our direction. and and there there a a new phrase came up in our world here today um that would at least were new to me which is public AI infrastructure. Um I was exposed to a couple projects. There's a there's a $400 million funded project on public AI. Um that was there were several >> Yeah. >> because ultimately the the challenge isn't so much the AI, it's the capital distribution out of the AI. >> Yeah. And into the AI, >> right? In both directions. Yeah. >> Yeah. Yeah. It's not whatever the the system in which it's an because I could write a beautiful science fiction story of a utopian ideal with uh with lots of AI that benefits everything. We could write a dystopian story to be the mirror and and it's not the AI that distinguishes them. >> Those darn humans. >> Yeah. >> Yeah. In all their forms. Well, I didn't mean to take us all the way to uh the end of this. I just >> we were we had a record short meeting. >> Well, I was gonna say when when someone brought that up, I thought, well, we haven't got to this topic. It was Yeah, we've been steeped in. But uh yeah, it was it was interesting the contrast of GDC, you know, two weeks ago and then and then this I was like, "Wow, both of them had some really strong themes that I think are are really, you know, good for us." Um >> yeah, >> even though there's still a smaller group of us that recognize that, but making progress. Yeah, Neil, definitely time for deep breath and a point. Yeah, that sounds good. All right. Well, um >> you can have a mimosa there, John. >> Have a morning. >> Yeah, >> that's it. That's it. It'll get you going. The interesting the fascinating thing about I only got this the invitation to come to this about a month ago and uh my sister lives here in in Washington DC and but they've been traveling a lot. They rent out their house and they actually make money renting out very attractive house in DC and they go, you know, get good Airbnb deals and they're really making money on it. So I pinged them. I said, "Oh, you know, are you going to happen to be in town?" Uh, you know, because I could stay with them. And uh my sister uh texted back and said, "Yeah, you know, that is my 65th birthday." And I said, she like, "We we're having a party." I'm like, "Oh, I'm coming for your 60th birthday party." >> Yeah. >> Well, enjoy it. Is that coming or >> No, that was last night. That was last night. It was fantastic. It was awesome. So, it was perfect timing and uh fly home tomorrow. So, >> all right. Thank you everyone and meeting well done. >> Thanks everyone. >> And uh back to the uh trust um steering wheel that for now on we have to talk about where the AI steering wheel. Okay. >> All right. Thanks, guys. Bye everyone. Bye. >> Thank you.