 I'm laughing a little bit because now I have to turn on four things at the start of every zoom call. I'm trying to do so automatically turns out to be problematic so hey, here we are. I was laughing because Dave asked if that baton was sharp. That's why I was laughing. You don't want to give Jerry anything sharp. No, it's not sharp. Don't run with scissors. Yeah, I'm of the school of try to reduce all the accidents before they happen if you can so. So for instance there's there's like a no fly zone over laptops, where, depending liquids are being passed around they can't fly over the keyboard. Okay, that just just trying to minimize the odds that something awful is going to happen. The audience spent got to go out to Camp David a couple of times with different kinds of retreats and she was working back in the White House, and she there's a big sign on the wall that has all the things that you're not allowed to do at the camp at Camp David. You can't race the golf the golf cards you can't play, you know, bumper car with a golf cards you can't you know, and it's like it is in the in the major who runs the place was saying yeah everyone of those has a story. You know, I bet I bet there's only a rule to not do bumper cars with the with the golf cards because somebody wants. Exactly there were there were once bumper car races. And all it does is inspire other people to want to do bumper cars, of course, but wouldn't have thought of it before but now it's there on the wall I mean hey. Good fucking long spring is arriving in Oregon very happy about that today's rainy but tomorrow goes to 70 something and then it gets really really nice so I think good weather is arrived here don't know about everywhere else. You're all feeling spring heats in San Diego where it's very spring like much at the time. springs is not so bad either. Yeah. Yeah, exactly. Cool. And I was just saying at the very top of the call that I was trying to come up with a series I was trying to nugget ties a narrative to explain neobucks as recorded videos and I started getting distracted by the many different things that thread together into into those but it's my intention to do some, some of what we talked about last time, which was hey neobucks needs an introduction and to produce those things. And I don't know exactly where we'd want to pick up on what we've got going so far. I don't know. I had a kind of a little I don't know if it's an epiphany exactly but an experience last week that I was just going to recount for you. Please. I'm talking with Michael Lenin might have talked about this a little bit last week too but it's still present with me. I talked with Michael Lenin about a session that's been going on in the GRC around a regenerative economics. So the, the, the Pam and Tunisia have been leading the session for over a year now so there's probably, I don't know, 50 or 60 episodes of the sessions, most of them are recorded. And Michael I think in particular has got some energy to do some kind of a public something with these things I mean that we started out talking about as a book. And I think Michael's been talking to you some Pete, I maybe. So you know this is all kind of all the threads are crossing a little bit, but I was struck by kind of two dimensions of it I mean I thought of the neobucks as kind of getting a book out to share information. I was struck that actually the process of creating the content is also a good way to activate people like in the GRC. And so we have a topic area this regenerative economics thing that a lot of people have been discussing from different perspectives. You know emotionally you could have like people writing chapters or something like that. But I was thinking would not gosh I mean you know somebody if we could, if people were motivated we could have like somebody running the podcast, and somebody putting out the newsletter and they don't even have to be all that tightly correlated really I mean they're, they might be pulling from the same kind of pattern and maybe we're, maybe there's a rhythm to it or maybe not you know somebody else is writing the book and there's chapters and the chapters stolen by the people are doing the podcast and they're, you know, I don't know, or, or chapters end up with research which is interviewing somebody for the podcast that shows up in the chapter and there's, there's, I don't know, kind of energy going around through the system, but maybe not all that highly coordinated, you know, you maybe you just have six different teams kind of around the same topic area. And then from my perspective you've energized the teams, you've probably explored the topic better, people probably learn, you know, there's a lot of in a GRC context it's just people learning, kind of for themselves and with each other. So the same thing why they were doing in the conversations. So, you could almost argue that the public, public release and bonus, you know, that's the, that's the, that's the, you know, if you get the value just out of the process and something. So I was going to try to encourage maybe something like that from this GRC effort. I don't know how closely correlated it should be with this group but anyway, so that was kind of a notion. So I think what you're describing is a really nice use case for Neo books in the way that we're talking about them makes total sense. And Pete did this kick up any thoughts for you about how we might go about doing something like this. Sounds really smart. I mean, at the current state what this means is GRC people learning how to put markdown documents on GitHub, I think, because that's practically the way we're doing Neo books. But other than that, it's like, come on down and play. I would say that a little differently. The workflow that we've gotten involves markdown documents on GitHub. And that's one way that people can participate but if they have already got a workflow they like Google Docs or sending stone tablets to each other or whatever we can work on adapting that to to what we're doing. Yeah, I had a, I had a conversation with Jordan this morning, which was on the police project but also the new book concept came up because we really need to use AI, you know, the, the, I mean, the topics we're working on. We need to have AI support to enhance to enhance this effort here. And when you when you look at how, how this technology is advancing my son, for example, you know, he works for same Sarah's head of corporate talent branding. They are actually training a chat GPT enterprise, this company that they're using a chat GDP GDP enterprise to train up for what are very technical issues, not because they're developing a knowledge base on basically linear optimization or movement and so on. And so the, the, the, the challenge right now really is to train chatbots to be specialists in a specific topic industry knowledge basis and what have you. So, and I come back to the Google team that trains an AI to win the against the gold master right so you have a team that was training this AI so we saw I'm my imagination has really been captured by training this AI so the entire Neo book that I've developed now face one and now face really is what this AI has come to think right because I started with the dawn of everything 10,000 years of of of agriculture the evolution of agriculture, the integration of food and agriculture into culture into the survival of civilizations way way way back. Now, up to the integration of social systems change management and all of that so this AI is now highly trained, you know, in, in looking at food issues from multiple perspectives. And so, and so we get we're already communicating through it right I mean we're running every contract every major discussions through the AI right now because it just blows up your your ability. It's an enhancement really that's that's not how we're looking at it, but it also turns out that the AI is completely dependent on who programmed it. Right so it's the the the integrity and quality of the programming team really determines the reliability of output from this AI. So if you are missing to frame it incorrectly, it may start hallucinating because it may come up with things that outside the periphery of of where you would want it contained. So that's where where where I'm really are now is I need to get out with this thing. I have, I have this sitting there. I'm working with farmers. And I'm working with people who are not going to go into GitHub. They need they need a different type of interface that is more user friendly, more practical. Yeah. And that's probably tool for a lot of other menus, not just farmers. So I'm sort of really aware we're going what what are we doing with this thing is, you know, and I really would like us to see what are we can be take this for a test ride. I mean, are we ready to move somewhere with this. Great questions class I think Pete and Pete and Stacy have questions and answers. So I hear you class. AI is an important power tool, I would say, so I don't disagree. I think I'm not sure. So, I teach people to use AI, LLMs and image generators. So I've had a fair amount of experience with people learning how to use LLMs learning, you know, learning that they're not very interested in LLM something like that. The way I look at AI in general is, it's a power tool. It helps you do stuff a lot faster a lot better. So if that stuff is text or if that stuff is images. It just kind of makes sense that you would do it. At the same time, I know there are people who for whom that, that decision seems very quick. And so they want to tip toe into using AI along with people. So I think that's one thing I think it's important to recognize that some people are ready to and some people aren't and also that I agree with you it's very important if you've got something that needs to be done quick teaching a bunch of people about regenerative ag or or whatever, social justice or whatever. You might as well use a power tool and instead of not using power tool. You focused in on specifically training or tuning fine tuning LLM to be to have a certain set of knowledge. That's, I think that's one way to use an AI and there are other ways to use it. You might use a general purpose AI and lots of different ways. You probably use both of them together at a fine tuned one and a general purpose one. So I think there's a larger, a larger talk there about what you would deal with LLM and why, why you would do it and have successfully expected to be it and you'd want to feed that process knowledge back into how you're using an AI. And I would love to have that conversation, probably not within the old books. Another, another thing you said at the very end is, can we just start doing something. I think to go back to what Dave said I think Dave actually had a pretty good example of what we could do right and it's and Dave I really like the way you sketched out a vision of actually doing something. Having it be multimodal. And maybe a key part is not to worry too much about how how that is coherent, you know, around a particular topic. I think that's a great approach. And I think, I think Neil book should think about that seriously and continue to kind of evolve that idea. The question for me, do you really quickly asked Pete what do you think I said it was a good idea. The second part of that answer is, how does, you know, you know so maybe kind of the way the way cloud set it. So when do we start, you know, who's, who's showing up every week or every other week or whatever rhythm it is to actually do that on some topic, right. So maybe it's a combination of the new books team and the GRC team or, or however it works out, you know, maybe it's a newly chartered team pulling people from new books and from GRC and from from other places. The AI on AI anthology team is another team that Michael Lennon and I share, and has been working on book publishing. And, you know, we're now we're wondering, how do we take paperback ebook. What's the next thing is it a video, or is it a podcast or how do we, how do we get people to read the book that we wrote, how do we do marketing around it. Anyway, I like where you're going with that Dave. And the, the next question is, has somebody raised their hand and said okay, this is, I'm going to take over the new books meeting and this is what we're doing next week we're working on blah. I raised my hand, and, you know, I'm going to have meetings on this day at this time and we're going to be doing that work. And the first, the first outputs are podcasts and, you know, blog posts will borrow new books is substack and use that or whatever it is. So, you know, let's do it. I won't be probably part of that I'm happy to help on tools and processes and stitching together other communities. But the content part of that project I probably won't be doing that I recognize the need for somebody be doing it. Stacy, do you want to, are you replying to cost or is it a different topic. No, same topic. Then, then I will go after you. Thank you. So I agree with everything that's been said. Yesterday when I was tossing and turning though I just want to add something I'm not negating anything that was said. When I was lying in bed and I was thinking, I don't want AI to tell me what to do. I want to tell it what to do. And towards the idea of what Dave's talking about what I mentioned Jerry to you Jerry and Pete before the call that idea to me that the way I saw that is the convening of people, but then people that are focused on AI would be there either watching the conversation and doing whatever it is they do that's from the AI side. There'd be other people interacting from whatever whatever they take what they need from the conversation. So, again, all everything that I've heard, all these ideas put together. So doing something. That's what I would like to see happen, because we already have our relationships. We're already in different places showing up for whatever reason. And sometimes it's a little more of one. Sometimes it's a little more of another. So I just I'll stop there. Thanks. Let me go back to class this question and see if I can maybe glue together some of what all of us have been saying a little bit cause it feels like you're saying hey, connecting up a eyes to bodies of work bodies of knowledge is essential and it's it's like a fast way forward. To me, I think some and I'm forgetting who what where but I think some of us are pretty close to Pete has been trying to make a GBT talk to my brain. That is down the road from what you're saying which is, Hey, here's a corpus of known texts. Let's let's use that as a corpus for a narrowly focused GBT. And I don't know which or who of us has actually done that already. Is that a is that a feature we've conquered already because I'm class you're busy sort of coaching GBT and asking your questions in a particular way, but have we built GBTs. This team, this team has not the. The idea of building a bot a chat bot with a corpus has been standardized. Anybody can do that with make with open AI, just by making a GPT Poe and flow GPT or our GPT competitors so you can do it with them too. And class you you've actually done that right you've created a GPT it's connected to that body work. There's a few of which I demonstrated a few times I have a personal advisor, which I get which is really very interesting. I've set up one for the produce project. So, so you so there is a very specific training but I have a foundation that I'm using in each for each GPT to lay in a framework, you know, so sort of a box of algorithms and stories and so on. The practical output here, for example, is we had a conversation about protein in crops right so to to create protein output from in the produce with a green course you know what types of crops. Could they call that content protein rich, and it takes this thing five seconds and you got a list of, you know, what you can call and what the protein content is and how you rotate them. So the, the, the, the research power and the knowledge base that is inherent in this thing right now is amazing. Right. And so if you can structure it in a way where a farmer feels competent and confident enough to go and just ask a simple question, you know, what what crop could I use for blah, blah, blah. I mean, that's amazing stuff now. And so the, because right now this is all time consuming but the interface would have to be engaging, simple, you know, maybe, maybe you can talk your question rather than typing it and those kinds of things. So that's sort of what I'm what I'm envisioning because I mean the these are all very smart people. They are actually I mean farmers are actually super engaged in applied technology. Right. I mean they're using GPT self driving tractors and what have you they're actually pretty good in deploying technology. Once they get the hang of it and get to understand it and I think that this tool is can be a game changer, you know, in the regenerative frame, particularly because the way we are thinking about connecting farmers with markets with CPG consumer package goods manufacturers. Right. So to to. So it's a very technical knowledge base but that goes further because I also want this this AI to think about the socio economic impacts of decisions. Right. So you want to be community specific you want to input the variables by community that make the AI aware of of things that no one right now even considers right. How many jobs do you gain and lose and how could you break this function down into more small scale businesses and today all kinds of opportunities available for that. So now going back to what they were saying a moment ago. I think the process Dave was describing of a community coming together to write pieces items nuggets or whatever about a particular topic, let's say regenerative economics and share them is is the process of learning about those issues improving those issues and basically creating a sense and and I think that just just like Neo books thinks of a book as a particular set of nuggets serialized in a particular order so that they look like a book. Similarly, a set of nuggets could act as a corpus for a GBT, but the creation of the corpus I think is really actually not only important in the sense of gosh we need to have some information to point this thing to. We need to have a social sense of community building trust building, and in the, the improvement of the work itself of the corpus. So what I'm trying to steer toward is, I think that this is, I'm really interested in the human and AI collaboration with things all of them together in some way that gives you class chat like access to really great bodies of work that can answer questions for people who just want to do that. But that also has the community side of people developing the corpus that the chat is based on and always improving it. And Dave I think sort of that that's a piece of what you were talking about is is like hey, we get we get lots of people interested in these topics. Some of them are just busy learning and ramping up to the topic. Others of them are actually improving the thing and making the corpus really good. And Pete the reason I was asking asking you who had sort of done this and class that that you have done this was that the GBT capacity becomes a sizable thing you do with any corpus that you've developed that could be any community's body of work. And that's how you sort of define the boundaries of what the GPT is good at it's like hey over here. There's a body of knowledge that we're thinking of as as a corpus for some reason, because it's vertically interesting and deep. Did that make sense. So, I'm trying to say both sides of this are really important the social human side as well as the GPT side. And there's a way of thinking about this that generalizes it so that aiming a GPT at a corpus becomes yet another manifestation of the work we're doing together to create bodies of knowledge. Yeah, I think, right, go ahead. Yeah, I mean, so, I guess I'm probably less confident still in the nuggets notion. I feel that I feel like the what I'm imagining is probably messier than that. It's, you know, there's there's not going to be a core repository necessarily. There'll be value in several different. So I'm probably organizing a thing in my head more like media channels. And in my framing of this I would say the AI is another media channels, or or what way to say it is a work team, right there's a podcast team and an AI team and a book team. They're working on the same general area, producing and learning and, you know, kind of co informed. But they have their own deliverable kind of, you know, so like there's one team that really wants to put out a book, but there's another team is really interested in the AI, you know, and those are that's great. So I'm kind of, I'm influenced by two things. One is just the frustration with all the channels, right and I feel like we can't choose a channel anymore you have to do all of them kind of right. And, and in particular trying to reach people with there's no one channel to reach anybody, you know I can't communicate with the GRC. There's no way to kind of get everybody's attention. In the same way Jerry it's like, you know, there's just, you know, there's no there's no way. So there's, there's, so my response has been well you may have to do all the channels. And so can you how do you create a structure that does all the channels well you do distribute the teams. And then the other piece of it that that I've been, I have been thinking for a long time that we wanted a regenerative media alliance, there's a lot of people out there producing stuff now. We want to aggregate those people so that we somehow amplify the effect. And I still wonder if you know books can have a similar kind of role there where you're you are the publisher that somehow making your authors more successful because somehow you aggregate publishers, you aggregate And so the regenerative media alliance notion was simply pushing out existing content. And I guess what I feel like the, this piece is kind of the bottom half of that pyramid or that time hourglass a little bit where it's like oh we're actually choosing a topic, and we can organize around and produce content that we're then going to push out in kind of the amplification, we still get some of the amplification amplification effect I think because the community is co promoting and co interacting and have shared networks or whatever, you know, and they're focused on this one topic and so that means that our public facing side will also be more successful. One of the interesting things before I go to Pete and class. One is fashion. The whole reason for nuggets is to create something that gets, that gets improved like Wikipedia gets improved so I have a riff on stocks versus flows and how everything is flows and we're drowning in the flows. So media streams or channels are flows of media and they just flows. To me, media streams are composed of nuggets because each post or each podcast is just a nugget, and the visible ones the ones that look like a web blog post or a podcast episode are mushrooms basically fruiting body out of this mycelial web of nuggets that are getting better and getting richer and metabolizing information. And I'm still not sure that I'm explaining this very well, but, but then the reason to go to nuggets is that it makes the production of things like podcasts and blog posts and sub stacks and all that it's relatively simple, and it lets them wire up to each other and get it get better over time in the same way that Wikipedia pages hopefully get better over time with good good participation and good editors. And so I would argue that dynamism is sorry I'll interrupt. I mean just, but the dynamism you're talking about is really critical I feel like that's been something I've not paid attention, you know, I thought the open was enough. It's not it's open and dynamic that matters right and dynamics the hard part, because that's for all the incentive structures fail. I think. So, so. I would think I'm wondering if maybe the nugget analysis it's a little bit scarcity based kind of like I'm kind of arguing and nugget will power because they're based. You're trying to be efficient. You're trying to have efficient allocation of knowledge by consolidating them into the nuggets so that easily redistributed. You'll probably you'll see you'll be able to maybe the AI organized the nuggets for us, I don't know, or you'll recognize a nugget business shows up in 40 different channels, it gets picked up by other people is it you know if it's important enough in the flow and idea will get repeated. So he gets repeated across you know four or five of the different products that's probably nugget. So he kind of discovered X post not free. And, and put into the flow so the flow that's the media channels are flowed but they're also stocks right I mean the podcast is a permanent repository. Only when it gets sort of stored someplace but you have a definition. Yeah, it's not it's not organized efficiently right that's the piece so if there's, but we have a lot of people doing a lot of things. I don't care, you know. I'm cool I think he doesn't have his hand up before class and then somehow the Q got flipped I'm not sure I purposely flipped it class started talking with my hand. That's what happened. So how would you like to run this class first. Yeah class first. Okay, good. Thanks. Yeah, so the conversation evolved since I put my hand up first. Let me try to catch up here. I think they for for your application, you want to set up a chat bot, and then got it jealously. Right. Don't don't have anybody mess with its assumptions and and and its parameters. Without you knowing about it and without having someone take a close look at whether that really is truly fitting in. So I'm using nuggets to basically evolve a story. Right. I mean, I sort of track what people are talking about. So for example, my last story was to soil, the role of soil microbes digging into soil health. Now, because I noticed that the conversations were drifting towards how does soil health with the health of the soil biome relate and link to your gut biome. And so I talked through this with the AI and and we ended up writing an article here. And that really resonated widely and it's already being picked up as yeah, we need to focus on nutrient measurements of crops that are linked to the nutrient density of the soil. So this is how markets evolve. So this is a freestanding nugget that may fit into a different conversation someplace else about the soil microbiome versus your gut microbiome. So that's the idea about nuggets. But the important part is that you need an interface between the GPT, the chat GPT itself and people using it and accessing it. If that makes sense. So you can have 10 teams working on the same chat GPT. And in fact, that's wonderful thing because you're training this thing to think far more partly and to look at what GRC is doing holistically as a whole. Because at the end of the day, everything that different groups within GRC are doing has the same underlying purpose, right? Regenerative transformation. So that's where I wanted to go. So the and then there is value in it. I mean, if you develop a chat GPT that is truly knowledgeable that dives really deep and produces, you know, create outputs to two questions that has value. And if you and the next step, what is really emerging right now is trust. Now, can I trust this source? So you are basically owning this chat GPT. If this thing starts screwing up, you are eroding your trust base, right? So you are responsible for hallucinations and what have you and avoid them. For example, I instructed my chat GPT to not respond on anything with a less reliability than 80%. Right? So the level of certainty has to be at least 80% before you use that. Now, and it will it will indicate to me that's a 95% and so so you're avoiding these these these pitfalls. So so I guess my key point is chat GPT is an asset. It's yours. You develop it. You own it and you're responsible for it. Thanks, Klaus. Go ahead, Pete. Dave, did you want to respond to that real quick? Yeah, I was a couple of a couple of things. So one is, I meant, I forgot to mention that we will like my region economics team will meet again on Friday. So we met last Friday, we'll meet again this Friday. So there is a little bit of continuing conversation. And yeah, I don't know. It's interesting because I kind of like again, there, I would have to imagine that there's somebody who wants to do the chat GPT in this milieu, who would take responsibility for it kind of like there is no, there's no nobody else right so if there were people who wanted to do that then I would say great let's go do it. I'm probably less focused on getting the right answer out of this than I am in having people explored and learn about it. So I'm not trying to be very prescriptive in the what. So I don't know. I'm not sure you're I'm going to take your wisdom about how to, you know, carefully guard the chat GPT and stuff like that. But I feel like really it's much more of people trying to get their wrap their minds around these kinds of things. I mean, like the podcast notion for me was podcasts are just great research tools. You get to go interview anybody you want, you know, so they would be a feeder kind of as much as they would be a product. You know, that's that kind of relationship we're looking for ways that we can understand. And the other, I guess one of the just will note that I did join Stacy. The importance of creating the shared language I think is one of the, you know, that's one of the things that could come out of this really is that we, we just don't know we don't know what each other's talking about yet but if you have a lot of these conversations the language, we get much stronger. Please. I wanted to. I really like Dave's, Dave's idea of different media streams and including a chat bot as one of the media streams or multiple chat bots different chat bots as different media streams. I also wanted to note that you can use AI and the production process as well as in the output process. So it's a power tool for research and thinking about stuff and creation of content and all that. So, when we say AI, you know, it's, it's different, different places in the Python. I also want to kind of, I want to continue to say that AI is super powerful. And I think people should use it, and I also want to be sensitive to people who have some who want to go more slowly in a more considered way. And maybe not even ever use an AI. I think, I think it's worth thinking through that. As I say that it also reminds me that another thing to consider when you're talking about AI is is your choice of model and how it was trained. So the big commercial ones that think really, they're really good at thinking, for instance, I like to use GPT for because it thinks better than the other ones. Cloud 3 Opus is supposed to be getting there. We'll see. A different consideration for using an AI, choosing a different model, you might want a model that hasn't been through the guardrail process or the, you know, the particular training process that one of the big commercial companies has used. If you want something more open source, or something where you're not shipping bits, shipping information back and forth between a commercial provider, maybe you want an open source LLM that's been more carefully trained in some more circumspect way and maybe you want to keep cell phones out rather than sharing all the chats with open AI or Anthropoc or Google or whoever Microsoft. I'm surprised how much we're talking about AI here. I'm interested. Again, I'm not sure that this is the place to talk about AI. I think that's a working group of new books, not, not the new books people. I also wanted to talk about. So now I'm going to talk more about AI. So having an interface to a GPT, or you talked about a particular particular set of personas that might use a GPT, say farmers. Farmers who are good at adopting technology when they need it and don't have time for crap technology or technology that's going to be too hard to, hard to adapt into their workflows. There's lots of different personas that might use. I, the, again, from my experience teaching people how to use LLMs. And then my experience watching other people build GPTs. I think we have, we have a long way to go to make them easy. And even chat GPT, which is, you know, tries to be so helpful and so easy to use, it takes a mindset change. I watch it happen. People come to chat GPT and they're go, okay, well it's got a text box. I know how to use text boxes because Google trained me how I type in, you know, sky blue, whatever, right. And they think of, they think of, they come in thinking chat GPT is an Oracle, like Google gives you the right information. For me, after having worked with it for a while chat GPT is, it's weird to say this to people who haven't used it for a while but chat GPT is a thinking partner and a motivational partner. It's an informational partner. And it's a way to shortcut longer processes like, like, I need it, I need three chapters on this topic go. And then it's, it's got it later, or I need a Python app that does this and this and this go. And instead of taking me an hour or two to do it, I get it done in three or four minutes, right. So, so I don't use it in Oracle mode very much I don't use chat bots in Oracle mode very much. And that's a big thing that you have to teach people who are used to search engines out of right. You also teach them how to deal with hallucinations. You know, I like to say that I have calling it hallucination or calling it lying. A lot of people say well the chat bot lied to me I don't know why it did that. It's like, dude, it doesn't have like agency. It's just trying to give you good answers. It's trying to have a good conversation with you. So the whole lying and hallucination thing is it comes stems largely from assuming that it's an Oracle that's always going to tell you the right thing. It's like that's, for me, that's the wrong way to use a chat bot a chat bot is a language power tool. It's not an Oracle. So, there's a, there's a big learning process I think, even when you tune up a GPT a chat bot a trained chat bot when you train a chat bot to talk to people about a specific topic or something like that or talk to them in specific ways. You still have to train the humans, you know, this is what this thing can do this is what it can't do. This is the way you have to approach it to get the best information out of it. It's parts of that into the chat bot and there's other parts you can't really. So, I think to kind of recap, when we talk about using AI, maybe we should maybe we shouldn't. I think that's an open question. I think that that AI shouldn't take over new books, I think it should be a media stream for new books, and just as important and not as and not overly important just like a podcast or just like a discussion forum I think a discussion forum is another media output and another media stream for what we're thinking of new books. So maybe, I think maybe we keep tripping over this new books thing right maybe that book thing makes it too much about. It's maybe too much of a noun, instead of a verb. What I think we're talking about is helping people think and learn. That's the idea that if we use nuggets and Dave I'm really intrigued by your, your analysis, economic analysis of why, why you might or might not use nuggets. But anyway, what we're new books is, I think it's a learning doing thinking process collaborative learning doing thinking process inspired by the idea of nuggets. And so then one more thing on AI. So I think I should stop there but I'm going to say one more thing about AI. Another thing about AI is you have to train people to use it. I see with my colleagues who are working on building training chatbots to do different things they have completely different approaches on how that works. Some people have a question and answer prompt style that they, they tell their GPT to use some people and and I can't express in a short frame. The depth of that. I've got a buddy who spends literally like days crafting a prompt one prompt days crafting a prompt of how to combine and synthesize information for a person who's coming to the chatbot. So when you come to one of his chain chatbots, and not even talking about the corpus just talking about the process of working with it. It's like a master's level course in whatever subject he's plugged in the knowledge base for right. It's not like a simple thing where I ask you a question it gives me an answer. It like goes through a whole process of helping you like a good professor would over the course of a semester. He's really literally packed a semester's worth of pedagogy into this prompt. And then hooked it up to knowledge right. So that's one approach that's not the only approach I think we're a long ways away from getting to the point where we have good GPTs and and well built GPTs. And, and also I want to recognize that it's not. We're not quite at the place where it's easy for everybody to come to a GPT and know what to do or how to do it or how to get the best results out of it right so we're working on that. So the technology, you know, AI technologists, essentially working on both ends of that problem. How do we help people use it better. How do we make the chatbots more more friendly more useful, more powerful over a longer, longer amount of collaboration with the user in collaboratively learning and thinking and doing a topic. I think there's a lot on the table and I'll say I don't know exactly how to catch you up so I'm going to let you float in the stream here with us. Send you on a witzel please. I just forgot the plan I was going to make. I was like, got involved in the scene to be a talk. You're the chat to give yourself a little demonic aids that I whatever I'm trying to do that I leave little hints in the chat that usually bought me back. I didn't do a good enough job with it. Like, again, I can read one thought on the on the nuggets is again just kind of the messiness of knowledge kind of that I just feel like we're trying to organize things that aren't Organizable really so you can organize them more I suppose and so there's I'm not against nuggets I just don't think I would want to to rely on oh I know what I was going to say. Thank you. So, and so to me it's like the nuggets aren't really the goal it's the, it's the learnings I kind of get back to learning so how do you enable burning amongst people. And part of it was I got to spend time with Bobby fish in this weekend talking about what he's been doing a crowd doing. And it's really amazing. I mean he's, he's kind of demonstrated the ability to recruit teams of volunteers and structure them in a way that they're really happy about being in that role they choose their own box as it were, and produce really valuable things. And that's kind of the inspiration a little bit for what I was thinking about around the region economics things could we create a space where people are doing the things they want to do in a structure that's really productive and constructive and rewarding. And they're volunteering, you know, and we could have infinite volunteers kind of I mean we're trying this is this is part of the abundance things I don't have to be efficient, because my pool of potential participants is, you know, global. But anyway, something you know I just feel like Bobby really has modeled some really creative stuff around, you know, finding people getting them into a place where they can participate, you know, helping them find work that's constructive it's really fascinating. And is it crowd doing dot world now is that his platform because all I had was a LinkedIn link. Yeah, you know, I don't I've been trying to get him to kind of like, you know, it's Bobby right so when you get hit with stuff you get hit hard. And, and it takes me multiple conversations to like, I don't understand what else going on. So don't think there is a place where you can really understand all of the different things. I mean, I can give you some examples of stuff but. But yeah, I don't I don't think there's a, there are no nuggets. There's just, there's just mouths. Got it. Where does that leave us. We've got another half hour right. Yeah, 35 minutes. I, I want to switch the topic a little bit. I really, really, really want to help new books, and I'm pretty sure I don't want to do new books. So, my guess is that I will, I will probably stop coming to the regular new books meetings. I'll probably call back for advice or not call back to a meeting just offline somehow. And I'm looking forward to working with you books on massive wiki publishing, whether or not it's massive wiki. Some of the podcast stuff, organizing information, the theory of nuggets, how, how to combine nuggets and things like that. I definitely want to help. I definitely kind of want to be involved in the periphery and I probably won't be coming to regular meetings. Yeah, so my question would be, what can I do, because I got this first volume it probably needs, I don't know formatting needs different pictures and stuff like this I mean how do I move on with this thing where do I go from here. I think another the next step. Well, so, without thinking without thinking yes or no about new books or something like that just if I were you what I would do is the next steps. The next steps, what I would do is get it out to 10 people, five or 10 people who really read the thing and take copious notes and say, class. This is a great book but here's some changes or class I would split this up into six books because there's so much information here or maybe maybe you need to write the guide to reading the bigger book or something like that right have some people really dig into it from an editorial content perspective and give you some feedbacks and move forward on that feedback. So, you may or may not decide to do that. It sounds like a pain in the ass, because it is. And, you know, maybe gets done maybe doesn't from a production point of view. The, you've already got it as a Google Doc right so Jordan and I when we did our book publishing project. We didn't experiment with Google Docs and Microsoft Word to do the print ready formatting. It was inconclusive we thought we thought that we were sad that the Google we tried Google Docs first Google Docs was we had some problems with the template they gave us. It didn't quite work the way we wanted it to. So Jordan tried the next round with Microsoft Word with the same template had the same problems. I think, you know, if you can read it in Google Docs and you want to get it out in the world, you go over to KDP or Lulu and give it I you go through the process of publishing it publishing the print edition is not that hard. And you just do it, you might have to go through a step where you take their template copy and paste your stuff into their template. Jordan and I found that that didn't quite work anyway so I think the what you want to end up with is a PDF basically see you print a PDF give it that and it's good to go. There's another step where you also have to do the ebook edition so the ebook edition, you simplify the formatting and give it the ebook stuff, but I that that would be the next, you know, editorial content. Maybe you want to do that maybe you don't and then publishing. It's easy to publish a step that that new books hasn't really figured out yet is marketing. So, you know, this is actually true of the anthology people to another another book publishing project I'm part of, you know, okay now it's in Kindle, or now it's in, you know, print on demand. And the anthology folks, we get to we got to go around for a day and saying, we're number six on the you know some incredibly narrow category that Amazon has, you know we're number six on AI, you know tools and practices or something And yet, I don't know how many sales we've had I'm sure it's in the, you know, handfuls of sales, not, you know, so, so then if you care, and you got a book. What do you do next. Kind of inspired by Dave's thing. The other thing that that new books is working towards and that if I were you I would work towards with or without new books is, how can I take this and repurpose this to different streams. How can I serialize this into substack. How can I have podcasts around this material. How can I give chunks of this material to different people to write their own book, or to stitch together, you know, hey, here's a chapter where here's a sub chapter here's a chunk of my book. Could you write a book around this. And, and just as you're doing also, I would have trained GPT's chain chain chat bots, ready for people to use, you know, to use the information, like I said, for you and me to use chat bots. It's not easy for everybody use chat bots. So there's another step of trying to make that either the chat bot so easy anybody can use it, or train your, your, your folks, whoever they are farmers or whoever train your folks to be able to use the GPT so there's that too. So there you go. That's all you have to do. So it's funny, Pete at the end of what you were just saying you crossed over into what I wanted to come in and say, and I'll start with class I may be over interpreting the way you started this call but when you're like hey this is all about the I was kind of thinking you were no longer interested in publishing books, and that you had corpus is you had bodies of work, which you were generating with GPT which you wanted to make accessible through GPT's meaning chat bots that know about this corpus and I didn't think you were still interested in actually publishing books, and the steps that Pete just described for publishing books are totally totally makes sense. But, but the reason for nuggetizing the reason for the approach I'm suggesting is that you can push a nugget out through our substack we created a substack so that you could actually publish a nugget and then get feedback because substack is a way of getting a text out in front of an audience and seeing what they say and getting comments back from them, because it has a comment system. The long conversation we had three or four calls ago about how to do commenting and whether we could have some architecture for getting, you know, community to talk around nuggets was exactly so that we can figure out how to improve those editing around them. Our efforts to make the massive wiki more of a wiki so that people can directly edit a page and offer their suggestions is also a way to make these nuggets sort of improve over time. But the way to the way to get the way to do a Tom Sawyer crowd painting of the editing tasks for a book or a new book is I think through community. And so a piece of what we were trying to do recently was, hey, let's, let's send some of these nuggets out through the substack as substack posts, which are the new web blog, I guess, and then see what happens, for example. So that is a way to do a piece of this. There are still some clumsy and awkward parts about like making it all sort of smell and look like a book and then making sure that there are no errors when it goes through the publishing process but but at the end of the day, Kindle Direct Publishing, KDP or Lulupress seem to be the two most interesting and most accessible ways to go create a book once it's once you're happy with it at that level. But the reason for nuggets is to have places where people can congregate and make something better. That's one of the big reasons for nuggets. And just to say the quiet part out loud, one of the reasons to have books is to market the market the fact that you have ideas around a topic. Less so than so that people have an artifact with which they can read like we did in the 1960s or 1970s. Bingo. And this is only called Neo books because Neo books are cultural artifacts that are widely recognized. And with the mycelial metaphor books are the fruiting bodies the mushrooms of the work together. But there are only one a podcast is equally a fruiting body or a mushroom, and chat access is a different form of mushroom. All the different manifestations of the ideas wind up being that the known cultural artifacts. What's the plot of anatomy of a fall. Guy falls out of a shell eggs in France and dies and his wife and mostly blind child have to go to court and figure out why. Thank you. There was an actress in the Oscars last night who acted in that movie and also is in the zone of interest. She got two nominations right. Yeah, two nominations quite interesting busy busy year for her. This is Sandra Hiller. Other thoughts this conversations provoking. My brain can't help but go back to the idea that that we need an ecosystem slash platform that speaks to this, all of this. And facilitates the understanding of what this is if if we want to do equal books more than just, pardon me, a new books as more than just ones these twosies. And that there is a way for people to wrap their heads around this. It's a complex enough idea that without putting some structure out there. The structure, even if it's just structuring a whole bunch of the bits and pieces that we've talked about, but assembling it into what is perceived as a platform. speaks to not just the, the creation of a book, but the creation of a community of people who want to learn and teach together. I think that I think that's well what we all want. And that I'm hearing us getting caught up in in the focus on the book and the focus on how we do the book. And not so much on the focus on how we do the community that would do these things, and how we would have the conversation that would lead up to holding a community that would do these things. For me it feels like we're, we're focusing on the nugget and the, and the technical bits and not thinking about the desire to teach and learn together and collaborate in a way that would, there would nurture solutions to these things. And that those, those solutions would emerge from the community. If there was a vision of what that community with that ecosystem that platform could be. So that's, that's my two cents. And Jose, I forget when you joined the call, because you might have missed Dave saying earlier, a piece of what you just said, which was what's really interesting here is a learning community. I'm really busy figuring out how these things fit together and making them better. And I'm totally agreed with that. And I think we have. This may only be visible to Pete and me I don't know but I think we have a minimum viable platform in massive wiki. And in our calls and in our channels we have for communication like matter most and the mailing list. We have marked on pages on GitHub are social objects that we can use and improve together. What we're missing is the community of collaborators who understand that and have jumped in the way Wikipedia has a giant community of two decades to build, you know, free encyclopedia for the world in lots of different languages. So we don't we don't have a functional community doing this, but I think we have the piece parts to do a minimalist version of the thing that we're looking to do. And when we're talking about adding conversational features or comment features or whatever else, those are merely to improve our ability to sort of do those that collaboration and have that conversation. Dave then Stacy. And so the way I was like Jerry was saying I'd be kind of imagining that we might have like a topic area and we'd have essentially multiple production teams, doing multiple products, if you will. So in the product that in class, we would like the product could be a chat to you can see but it could be a book it could be a podcast series. But they're all kind of around the same nuggets, if you will. And then there's a question of how organized the nuggets are. I've ended up imagining this is to me this is one of the problems is that the coordinating is to I don't know how Wikipedia did it. And we've not seen many more Wikipedia's. So it might have just been a lightning strike you will never going to be able to repeat it. I don't know. But looking at my world now, and like, or, you know, and like, I don't think everybody's going to come to this space Jerry and use the matter most and, you know, use it right I'm going to have to go to where they are. So I mean, like I have some people who are paying attention to the GRC their meeting together. You know, if we were to spin off these product teams, those product teams are going to define their own platform for the most part, I think, you know, like about the products of podcasts, they're going to come up with their own tools and they're going to coordinate around a podcast, and somehow or another the problem will be the cross filtering back into the rest of the community right so that there is some kind of, you know, knowledge spinning through the process. So that's a knowledge is getting better the learning is getting better. But there's a weaving component I think that might be essential. And I do like the idea of trying to, it's like, look, let's use like the GitHub as one of the shared resources if we're disciplined at that, that will help our weaving. And then you have to nag people at time to get them to do that and they'll do it 20% of the time. Right. I mean, I don't know. That's my experience with the world right now. And so, you know, this notion that we're going to get it all organized it's going to be tight just doesn't work anymore I don't think, and, you know, and I don't think we're going to build another Wikipedia either. I think we have to figure out ways that are much more kind of spontaneous and piecemeal and random to do this stuff. And just very briefly, I don't know that this is going to be tidy I don't know that we're proposing that it is tidy. I think that the concepts we're talking about here are a way of converging ideas rather than letting them just splatter on the wall. And convergence doesn't mean that they end up in neat blocks it just means that there's a sense of orientation there's a sense of context, there's a sense of collaboration in the medium sometimes go ahead Stacy. So what hardens me when I listen to Dave and Jose speak is that they're addressing something that I think gets the focus of the importance to go from the ground up gets lost in a lot of the approaches that I hear when they come from very from the technical community. So it's not a purposeful thing. It's just happens that you know there's like a goal and you're trying to do something, but we forget isn't the whole point to to bring the bottom up and empower that, and that gets lost in some of the things other than what I hear from Dave and Jose, and I'm, I'm trying to inject. So, you know, in class you're right there in the middle, you're right there. Thanks. Yeah, one, one thing that I could envision is and we talked about this before is to have a platform that gives you a general entry to new folks, and then gives you the choice to follow specific topics. Which may be individual books or it may be a series of books within one topic, you know, like nature based solutions may have gotten world and others NGRC and other things in it. And then give us the ability to use this as a link to go elsewhere with right so we don't have to build three or four platforms. But if you build an OGM platform for new books, you know that that can then be accessed through my website or through other places right now we have the credit that OGM deserves for creating this whole thing. You know you're creating an access platform. You can do all kinds of things with it. So that's, I think, is what we, Jose, and we talked about this before. If I may go, then we can also play with it because if you are building rooms, right, I mean you're building a platform with multiple rooms, then Jose can do something different in his room than I would do in my room. So you can customize it to my specific audience and for others. I like the question do we have a platform. And I like the first part of your answer Jerry, and I'm kind of meh about the second second one, even though that's a surprising thing for me to say. The first thing is like we have a platform we have these calls. So, to have a platform. We need a social platform we need the ability for us to interact with each other and then we need an external sociality to we need to, you know, mix with the outside world, have people know what we're talking about teach people how to think whatever it is that new books does right so so kind of let me come back to the way I started out the most important platform is the social platform and then in the platforms we have internal platform and external platform. So then let me switch from social platforms to technical platforms. Jerry I kind of like and and maybe don't like your answer about the technical platform we have. It is the one champion my me so I really love that part. I also think it's the best, I wouldn't be championing it because it wasn't the best I think it's literally the best. However, I also know that people want to choose different different technical platforms. And in some cases in some use cases or or socialities or whatever, even the massive wiki platform is not the best you know it's best for them to use blah. So the platforms I can think of real quick. There's massive wiki ish one markdown, especially markdown is actually the important part of markdown and files, and a way to version the files. And if you start talking about ways to version files get it's going to be kind of the winner out of that, even though there's other other ways you might choose. The massive wiki platform I think is the best. Another one that's very commonly used is Google Docs and Google Sheets. So Dave that's probably product, what are many of those teams the podcast team or whatever are going to say well, we all know how to use Google Docs we all, you know, and pretty quickly the whole group is either better or worse organized. They hopefully have a shared Docs folder they start keeping track of stuff and Google Sheets, they have a bunch of written stuff and Google Docs. The whole thing is really hard to find stuff in and it's always confusing and messy. And that's the best case the worst case is you've got everybody's got their own Google Docs and you share individual docs back and forth and it turns into a big mess. I'm not, I'm not trying to make that sound bad. I think that's a certainly a valid way to do. It trades the, it trades off in a bad way I think the, the using Google Docs as the tech platform Google Sheets and Google Docs using that as the tech platform trades quickness and familiarity of starting versus utility and flexibility over the long term. So massive wiki is a little bit hard to get started with, and then it just gives you back dividends of usability and flexibility and power and things like that. Google Docs is kind of the opposite way it's like super easy to get started because almost everyone knows how to use it and everyone's got a Google account except then a month or two from now or three months from now the whole thing has gotten air shaped and it's really hard to find anything just different, different styles notion. If you can get everybody to agree to use notion instead of mass wiki or instead of the Google Docs thing that's another great one. Miro for some, some people some teams that's that's that's what people use. People even go out into like there's teams that just do great with a discord, you know they need a chat thing and everybody manage there's their own docs but you can touch stuff into the chat. Some communities are superb at using discord for that. Discourse would be another platform that you could use it takes a little while to adopt it, but it's super, you know, super transparent and I use it and, and builds a great knowledge repository so if you're building a knowledge community. You might bias towards discourse rather than discord, for instance. All right, long, long story short. I do think massive wiki is one of the very best things and especially continues to give back in the in the out months and the out years and it continues to be collaborative and interoperative with many so many other things that you win, I think, including AI actually. So, and then lastly, I would love to talk, I have a lot of interest and passion about how Wikipedia got to where it is that it works and and how, you know, where how it got to that thing. Another, another existence proof of something like Wikipedia. That's not quite as big but is is very, very successful another existence proof is wiki how wiki how it's an amazing kind of. Start and continuing to cultivate story that is actually a nice counterpoint to Wikipedia. Wikipedia is an amazing and wonderful resource. But some of the ways it ended up growing it got overly bureaucratic and wiki how has a little bit different philosophy so it's going slower and it is smaller now, but it's also got a good collaborative community that works within it. There you go. Thanks, Pete. Rick. Yeah, thank you. Just some reflections based upon what people have said here, Jose, I thought what you were what I captured from what you were saying was, well, how can we make learning more dynamic, more enriching flourishing what people want to come and learn. You know, it's, it's, it's the old intrinsic motivation of how people, you know, develop inquiry skills curiosity, creative co creativity, whatever. And, you know, Pete what you were describing for me was, you know, there's so much out there and the question is how, how can we create sort of a hybrid of systems that allow asynchronous synchronous communication where people can organically grow and develop interest and whatever. And class I read your, your GM thing on nuggets. And I was thinking, well, you know, that is something that could be put on some place where people can come and respond to that and my response which I haven't, you know, come up with a bit be a nice place is where, you know, I think there's a typology of nuggets and I think it's a bit of a broad term. And so you can have, you know, nuggets that are focusing predominantly on content, you know, the science of things. Then you can have nuggets predominantly focusing on process which is the learning process, which enables the content we have plenty of knowledge of what we need to do what we're lousy at the learning process to actually put the content into it. So you think about all the sort of, you know, cultural social barriers for, you know, elevating regenerative economics, agriculture, etc. I mean, it's, it's, there's so much, there's so many barriers there and really comes down to how can we captivate our, our collective wisdom to take on the meta crisis, because unless we learn how to deal with a meta crisis, then we can't deal with the poly crisis of wicked problems. So how can you create this sort of dynamic learning community that would be intergenerational at different levels. I mean, that's the sort of, I mean, it talks about completely transforming the future of learning. And so I'm, I'm sort of wanting to sort of go more into the actual doing of this. And so if you were to take, for example, classes, you know, posting on nuggets and put it somewhere and see if we can get a nugget conversation going about how nuggets can be used in different ways. I mean, it's just, you know, we spoke about this two weeks or I couldn't make it last week, and that was how to, how we, how we, how can we practice what we're preaching, basically, and learn how to do it iteratively, organically, emergently, yadda. So I think we need a little, a little, you know, Jerry, you mentioned we got to start somewhere. Well, let's, let's focus on one small piece and just experiment amongst ourselves using somebody's contribution and and seeing where that will lead to. So, interesting enough, I just wrote a blog post on sacred silence and borrowing from indigenous wisdom. And I'll share that that blog post because I think unless we, we enable sacred silence, we're so trapped up in the, the, you know, the marketing fast paced world, shining object world that we're so distracted that we can't really do any deep work, deep learning, etc. So on that note, I will remain silent. Thanks. Thanks, Rick. I wanted to note that Jordan Sukut is working really hard on solving the meta crisis and building a platform that that enables people to learn and grow and do and be together. So, Jordan's really doing that. We're really dedicated to doing that remains to be seen whether or not he'll be successful, but I wish him the best. Could you say what it was I couldn't. Could you say Jordan Sukut is the person and Lion's Berg is the name of the platform he's working with. Do you have a link or something that we can look at. Lion's Berg dot wiki I'll put it in. Okay, great. Thank you. And it's it may be more approachable. If I could introduce you to Jordan. Cool. Since Pete brought that up, does he still have that page that I think forest had put up of how you could apply for like if you have a proposal it was a very simple way to put forward a project proposal for funding is that still there. I don't know if they're doing grant making right now, even. And I don't know where a page like that is there might be one on on the wiki, but he's started today actually restart Jordan continues to kind of restart every every once in a while. Because he's not got the the right the an approach that's getting enough traction. You restarted today on something called critical path conversations or something like that. Thanks. It's going well I think joins join is continuing to go do great stuff. He's also massive wikis most avid user. By, by a league. Because it's the best not the amount of the amount of pros he's put on the Lionsberg Ricky is enormous. He's got about 1.3 million words and 15 like a dozen wiki books he calls them. So much work. I introduced Jordan to my partner. Because we have, we would like to go into the police life, but we need funding obviously and. And so, Jordan has a really well done organization and place and see where that goes. Well, we are at the top of the hour at the end of our call ish unless there's something anybody would like to add to it. I think you may have missed feel like savings time. You're muted. So we're not hearing you. Well, I'm sorry about that. Is it I thought it was was it 130 or 230 I'm confused now you completely confused me. So these calls started 1030 Pacific every Monday. Yeah, 230. Okay. I feel like that's 130 Eastern. Okay. All right. Okay, that's I didn't change my calendar. That's what it was. Okay. All right. Exactly. I was I was noting that you'd come in kind of late and I hadn't put together there was like an hour gap there. But yes, we we had already been going for about an hour when you when you did join us. So, so we're ready to wrap. All right, that's fine. Good. Thanks, everybody.