Mere Morpheus

Mere Morpheus artwork

DeAI Summer Has Arrived | TEE Rollout, Model Wars & One-Click Agents

49 minAudio

Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via their podcast app and to have them all in one place. If you'd like to participate they are held weekly on the @MorpheusAIs Twitter spaces. So enjoy Episode #76! Recorded: 4 June 2026 AI SUMMARY: This week's DeAI Summer edition highlights Morpheus' record single‑day throughput of 400M tokens with 72% API gateway share of top models GLM 5.1 and DeepSeek V4. It covers accelerating demand, the provider flywheel and usage differences between API and peer‑to‑peer. The rollout of TEE‑enabled private inference on Morpheus is detailed, including live models (DeepSeek 70B, Gemma 3 4B, a Qwen variant) and upcoming "max privacy" filtering. Model comparisons note the trend toward smaller distilled models alongside TEE assurances for sensitive workloads. A frictionless agent deployment with Manifest + Privy is previewed. The discussion touches on PewDiePie's Odysseus, the surge of local AI, institutional voices on AI and why real utility is driving a DeAI Summer decoupling in market performance (Morpheus, Venice, NEAR). It closes on X402's hype‑to‑usage cycle and the call to test TEEs and new models. Timeline: (00:00:00) Intro (00:01:23) Morpheus usage metrics surge and API vs marketplace mix (00:05:37) Introducing TEEs on Morpheus: why privacy matters (00:11:47) TEE models live now and easier discovery coming (00:15:11) New model releases and the small-model trend (00:19:03) Model popularity on Morpheus: GLM 5.1 vs DeepSeek v4 (00:22:02) Token spending habits and enterprise optimisation (00:25:37) PewDiePie’s Odysseus and the rise of local agents (00:28:51) Cultural signals: creators and religious institutions weigh in on AI (00:35:05) Manifest x Morpheus: one-click web agents and easy onboarding (00:39:54) DeAI Summer: real utility decoupling from macro markets (00:44:21) Agent payments x402: hype, trough, and path to PMF (00:47:17) Community call-to-action: try tools, join Discord, build If you would like to support me for the value I provide you can use the referral link below when depositing capital and we will both earn bonus MOR. Alternatively you could send me a token of appreciation directly to my wallet ;) Referral Link: https://dashboard.mor.org/capital?referrer=0xE935f231c99c04Ee0b4532a3d0BdA81B152a0384 Wallet Address: 0xE935f231c99c04Ee0b4532a3d0BdA81B152a0384 Connect with Mere Mortals: Website: https://www.meremortalspodcasts.com/ Discord: https://discord.gg/dX9BEhT3kz Twitter/X: https://twitter.com/meremortalspods Instagram: https://www.instagram.com/meremortalspodcasts/ TikTok: https://www.tiktok.com/@meremortalspodcasts Value 4 Value Support: Website: https://www.meremortalspodcasts.com/support Paypal: https://www.paypal.com/paypalme/meremortalspodcast

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

David JohnstonGood morning and happy DeFi summer, DAI summer edition to everybody. Hey, Bo. Hey, Kyle. Hey, Thomas. How's it going, guys?

Thomas BorrelHello. Hello. Doing great.

Bowen BelmerHey. What's going on, everybody?

David JohnstonDoing great, man. It's a it's a hell of a week. You know? It seems like everything decentralized AI is taking off. Usage on Morpheus just went nuts the last week. I counted a 400% increase in the last ten days, which is just incredible. So, yeah, seems like, things are really catching on. But so should be a good, discussion today.

Bowen BelmerYeah. Absolutely. We have a handful of topics to talk about. Do you want me to dive right into it or should I give it a second here?

David JohnstonYeah. It seems like people are piling in. So why don't we go ahead and kick off? Bo, I think your volume's a little little quiet. You get a little closer to the mic. Don't don't be afraid to Oh, is it okay? Dulcet

Thomas Borreltones. It's good. Yes. Sound sounds good for me. Okay. That sounds good for me too. Although you, David, sound like too far away. So Interesting.

David JohnstonInverse speech. Oh, Oh, no. You go ahead, Bo. Now you sound good.

Bowen BelmerOh, okay. Perfect. Alright, everybody. Well, David hit the lead story for this week right on the head, and it is usage statistics for Morpheus as a network and as a inference marketplace. So if you visit tech.more.org, and then on the land down that says, is Morpheus growing network throughput and session stats? You can see all of the numbers that I'm about to be speak about here. And

over the last few days, we surpassed 400,000,000 total tokens processed on a single day, which was, I mean, put it this way. This chart is up and to the right, which is exactly what we wanna see. And, you know, we hope to see that continue, but it's really really awesome to see that the usage is growing. You can see what the top models over the past thirty days have been. As of right now, it's GLM 5.1 and DeepSeek v four, as well as GLM five are kind of the leaders in terms of number of tokens

put through, and also in terms of number of sessions. Another stat though that is very interesting is the share of the throughput that is happening through the API gateway. So as of right now, it is approximately 72%, 72.4%, which is, of course, a high number. But over time, you know, as as the marketplace matures, we suspect that that number will probably go down and level out somewhere along the way.

And meaning that we'll have more people who are not we have a ton of people using the API, of course, which is a great thing, and I wanna shout out to everyone working on the API. That is actually how I myself experience Morpheus is is via the API. But if that number continues to trickle down over time, that implies that more and more people are tapping into the marketplace. Anybody takeaways or anything

Kyle (Bow Tied Blue Fin)to the usage and You're cutting in and out a little bit, Bo. But No. I just So yeah. No. It it's great news. It's overall, I think the percentage of people who are using the APIs is definitely going to fluctuate. Right? It's the API is one light client on the Morpheus infrastructure. So whether people are choosing to use that light client, whether people are choosing to use another light client, whether people are choosing to use,

to create a node directly, we love all of it. Right? So it's just in for for us and for the things that we're working on developing, we just want to kind of incentivize and and help drive demand any way that it's coming into the marketplace. So to see direct demand, it's amazing. Right? The the more demand that comes in, the more usage of these models, the more incentive there is for providers to come in as well and,

you know, monetize some of their excess capacity or or their, you know, DM or or however or or extra open anchorage, whatever it may be, and, you know, monetize that and provide that. And the more people who are in and using, just the stronger the system's going to get. And the way that we see it, and I've said this before, is that there's a little bit of a flywheel here. Right? More demand is more opportunity for providers. More opportunity providers means, you know, more

availability for for the consumers and and more, you know, trust in in the depth of the system and and both really help increase the value proposition of the Morpheus marketplace alongside each other. So really great things, really, really, you know, exciting to to see this, and we wanna take this as kind of a jumping off point and and continue to kinda get the message out, get the word out, and and continue to drive usage.

Bowen BelmerCouldn't agree more, Kyle. And this actually, Kyle, while while you're on stage and unmuted brings us into the second topic that we wanted to touch on this week, which is, of course, the fact that that is is rolling out and has rolled out, within Morpheus. Can you speak to that at all? I know, you know, we're as a community, by the way, just generally, this is, in my opinion, and I think in the opinion of others, one of the biggest things we should be talking about

at this point. It is one of the biggest differentiators in terms of, like like, product offering that you can't get other places. Or if you can, you're not getting it in the same decentralized permissionless way that you're getting it with Morpheus. And so, Kyle, do we have any update on on the t or any just, like, use of stats or anything like that that's worth mentioning this week?

Kyle (Bow Tied Blue Fin)Sorry. I was on mute. We have Alan in here who who can give some more of a technical update, but I'd say, overall, you guys probably saw, some of the git commits come out that the, main branch of the Morpheus router was updated to further some improvements and some compatibility on the TEE side with Secret Labs. You'll see more commits

coming in there to, you know, over time to look at additional providers and additional way to kind of not strengthen the the TEE structure, but expand it for more availability. We are in the process of, working on some documentation and some more marketing style information to show, like, what what is TEE? I mean, most of us here know, but the larger population doesn't. Why is it important? Why is this valuable? And why is this kind of such an amazing launch,

that we've had? Now, the the next big question for people is, well, how how do I use it? Right? And and that's what we're going to be explaining. Well, there's, similarly to the API. Right? There are other, like, clients that have been developed that kind of simplify the facilitation of creating a node, running TE, you know, end to end private inference, you know, yourself or for your business or or whatever or for your agent or whatever it may be. So,

a lot of great stuff in the pipeline informationally that we're going to get out, and, you know, technically, that that we're going to get out as well. So really excited about it and, you know, more coming your way.

Bowen BelmerPerfect. Thank you, Kyle. I did invite

Kyle (Bow Tied Blue Fin)Oh, is there anything specific on that that maybe you were confused about or you think would be kind of valuable for me to talk about with everyone? Or or if you want Alan to jump in on on any of the technical details, those, of course, are all usually more deep in in documentation, but we you know, Alan can do some high level stuff if you'd like.

Bowen BelmerYeah. I mean, I didn't have anything really specific. Although, yeah, Alan, I I did send you an invite to jump up and speak, if you're able and and want to do so. Nothing, like, extremely specific other than I I do agree with you that there's kind of two types of people, the ones who are, like, more deeply tech savvy who are gonna wanna see the full docs and, you know, essentially, the proofs on that it works and and how it works. And then also just from the user's perspective, like, how in practical terms do I start using these TEE models, as opposed to the non TEE models? And should I be using them, you know, if if you're any given user?

Yeah. I I guess I don't have any, like, super specific questions other than, you know, I know things are in the works to clarify the information for both of those types of people, which is, you know, great and and all that we can ask for.

Kyle (Bow Tied Blue Fin)Yeah. So let let's save the the technical stuff for a few for for those documents and and for future conversation here. I know we did kind of a session on that maybe a month or so ago. Probably time for another session, but let let's make that one a little more structured. So I guess, just to give a little insight on on your point on, like, what when to use t and and how to use it, I would say that everything

right. The market right now has moved to let's use a sledgehammer for everything. And I think that we need to be a little bit more precise in what we do. Right? You see that date with Everclaw, implemented, like, some model routing where you're not using, you know, the highest model for for every type of action there. I think that privacy is the same way. I think that there are things that we need to use TEE for. Right? There are also drawbacks to TEE being, you know, potentially a little bit slower

prompting times, higher cost. It's not necessary for asking the weather. Right? We all know that meme of, the guy, you know, asking the weather with with Claude Opus and and kind of just using the sledgehammer. Like, that's not necessary. So I think we need to as we start to use these new tools and these massive models, we need to choose when what are the right applications for the most privacy? What are the right applications for, you know, the medium level privacy, let's say, by going to know directly, and then what are the applications that you could just hit the API

with with the, you know, an OpenAI compatible endpoint for for a simple call. So I think there's we all have to figure out as an industry, you know, what what's the best tool for the job in different places.

Alan (Nomadic Rogues)I I agree. This is Alan. Just figured I'd jump in a little bit. One of the most difficult things we found, especially for consumer, side of things was, well, which model is or isn't team enabled? And it's funny you think about some things from a tech standpoint so long you forget or you just you you don't feel the pain on on the the user and the consumer side of things. So just for awareness, we've got three on main net. There are three posted t models and they do have a suffix called quote, you know, colon t e e,

deep seq 70 b, Gemma three four b, and I think a quen model as well as out there and they're all they're all suffixed with t e e. One of the things that we found is, again, from a client standpoint, it was very difficult for the end user to, you know, figure out which one that is, you know. Again, to Kyle's point of you gotta find the right use case. What really don't you wanna let out of the box? Is it that note to your boss?

Is it that financial statement? Is it that health question? You know, those are the kind of things that like, look, if you don't want your prompts to leak anywhere, this is when you would use the TEE stuff and either individually or from an enterprise standpoint is is when you don't want any leaks. You don't want to use have your data be used for the someone else's next training set. So coming up soon in the we're in testing right now, the client that we bake in for the the Morpheus Lumin node will have a switch that says, hey, show me the high privacy ones. Show me the show me the max privacy mode models that are out there.

And right now, we're limited to what secret offers, which are those three models. And hopefully, we expand and bring on other other providers and work with other folks that have a good tea background, kinda like secret labs, those offerings will get better and better. And so a, there's making it work and making sure that it's, you know, not just promises but physics. B is, can I find them? And, you know, c is what can I use them for? So

and then d is let's get more of them so they've got a a better a better marketplace for that. So all good things on the TEE side. If everybody followed those suggested on going to tech.more.org, There's also the t road map in that same site. You can kinda check-in. Last time we updated that, I think it was about a month ago. That'll be worth another update in the next next couple of weeks.

David JohnstonAwesome. Appreciate that update, Alan. Yeah. I mean, just to put it in perspective for folks, like, all this different work on privacy or the things people are building that make it easier to use is having a real effect. I mean, the usage in the last four days was equivalent to all of the usage in January. Like, that's that's the scale at which it's it's growing. And I was running out some models. You know? At this rate, you know, there's just gonna be a ton of usage by by later this year. So, you know, sort of the cumulative efforts of of everybody are really having a a pretty amazing effect. So it's it's cool to see, you know, we're starting to to see that product market fit

start to hit. Absolutely. Yep. Think Bo might have dropped off, but there's a there's a few other aspects we wanna talk about this week. A couple of big new models. I don't know if you guys have had the chance to play with any of them, but you've got m three. You've got a new Gemma model that Google released specifically for agents. I think it's a 12,000,000,000 parameter. In my experience, they haven't quite been as good as the larger models, but the the gap really seems to be

continuing to close. And so we could see more and more of those smaller models reaching the frontier level. Thomas, I don't know if you had any time to to play with the various models that have come out. Curious about your testing or if you just want to comment on sort of where the trend's going.

Thomas BorrelYeah, overall, I would say I've had the same output as you've had, particularly with GEMA. Not as impressive as the GLMs or even Kyma K26. But I think overall that distillation of models into smaller versions, smaller, most specialized version, an interesting trend, right, because you had sort of the extremely large models, then we went to a mixture of experts.

So it's still big, it still needs a large cluster of GPUs to run, but you get those experts that you can spin up that will utilize smaller resources. And now a better distribution of knowledge across distilled models that can run either locally on a phone, locally on a laptop, or in the larger data centers with a bunch of TPUs. So

interesting trend, right? We've all seen that through compute with centralized, decentralized mainframe, distributed computers back to cloud. And I think we're sort of seeing that with the intelligence, with everything being jam packed into a model, then a mixture of experts, then distilled models that go towards the edge. But, you know, overall, right, with all these models, what we see is you can run certain things locally,

which is great. I think it's a very important part of protecting your data, retaining ownership of your data, retaining sovereignty over your data. But then there's always those use cases where you have to go with the larger models, and this is where implementations like what Alan and Kyle have done with TEE is crucial because now I have those guarantees that,

like, I can say with certainty which version of model I've been using, which GPUs I've been using, and that my data has remained secure, which, you know, the moment I have either sensitive data or I have some obligations as a company, some fiduciary duties as a company to protect the data of my clients, having that kind of certainty is crucial. So

long rent, but that's where my head's at with those smaller distributed models.

Bowen BelmerAwesome. Appreciate the the thoughts on that. Beau, welcome back, my friend. I think I'm back. Can you hear me now? Yeah. Loud and clear, man. Okay. Awesome. I am personally victimized by x spaces more than anybody. I swear. No. I'm just kidding. Thank you, guys. Were you were you talking about one of the new model launches by chance? Yeah. The the new Gemma models,

David Johnstonm three, you know, continues to be a bought a lot of releases, though. At least from my testing, GLM 5.1 continues to be the best model. And if you look on the usage on Morpheus, on the API, and it's interesting. On the API, which you see is the blue bar on the chart, if you go to tech.mor.org, GLM 5.1 is the number one model. But if you look at the peer to peer usage, most of it's coming from d c from DeepSeek v four,

which is interesting. Something like 80% of the DeepSeek that gets provided is through peer to peer. And so I wonder if that's what's driving this increase in demand that we're seeing on the peer to peer side is it looks like it's mostly people consuming DeepSeek v four, which is which is interesting. I I I thought it was a really good model. I tested it out a bit when it came out, but I wonder, you know, if it's all the events in Malaysia, the work that Renee and others and Eric have done,

you know, educating people about

Bowen BelmerMorpheus over there. I wonder if that's their default that they're going to. I was wondering the same thing, actually, if if that was, largely Morpheus Asia driven, which would be awesome. Yeah. Yeah. Well so okay. I did have a question. I don't wanna make us go back to TEE, but I did have a question. I was I was trying to ask it before, but I think I was not I I wasn't coming through. When Kyle was talking, he had mentioned Morpheus skill, David, and that

is can Morpheus skill determine or make suggestions right now as to what models should be used for what? Can you touch on that? Sure. There is a built in skill as part of, Morpheus skill that,

David Johnstonlooks at the, request that's being made and can help select a model, based on what it needs to to get it done. So if you're just asking a, you know, basic information question, you can probably get it knocked out with, you know, g l m five. Right? If you want a coding task, it's probably g l m 5.1. Right? It'll go more towards, a frontier model based on which models you have enabled. So, yeah, I mean, for me, that's cut a lot of my token usage is,

as I said earlier, not using the sledgehammer on every request. But, you know, to be honest, like, a lot of times, I use the sledgehammer.

Bowen BelmerYeah. I was gonna say I'm so guilty of this. Even if I'm just Yes. Like, writing, you know, a message, like, having a draft message for me, and my my mind instantly goes to, want extra good message, so use extra powerful model. It's like, it's not necessary. But for some reason, it's just I think it's just human nature to be like, oh, big shiny button. Like, we should use that one. Yeah. Well and, you know, I think it's okay. Right? Like, you know, these these models

David Johnstonare great, and, you know, why would I want a, you know, a c plus answer when I can get an a plus answer? Right? And so it's like, you know, I, you know, I I I think I think in the personal context, it's fine. I think where the optimization really makes sense is for companies and businesses that are spending huge amounts of money. Maybe that's that's the other trend we should talk about this week is people have finally started getting the sticker shock for their their API bill. And if you've seen, like, the memes or videos, but they're really starting to hit where people are like, wait a minute. What's this million dollar API

bill that I just got? Like, oh, that's all the tokens we told our employees that they had to spend or they were gonna lose their jobs. Right? And so I think it's just like the next step of maturity is, like, one, everybody wanted to use that tool and then realizing it has a cost and having to step back and say, like, okay. I just don't wanna be token maxing. I want outcomes.

Kyle (Bow Tied Blue Fin)Yeah. That that's the point that that's kind of the target user that I was thinking about there. Right? Like, individuals yeah. We're we're cranking tokens. Right? But asking, you know, us sending a 100 prompts per day even is, like, insignificant when you think about the grand scheme of things of either AI coding or or workflows being run on, you know, internal data or internal operations. You know, that's that's where you're going from,

you know, a thousand dollars a day usage to 200 or or less dollars a day usage. You know, us, you know, spending $4 or or $7 is, I think, less impactful than when you're managing, like, you know, your operating income and and so on. So I I think at scale, it it becomes a much more significant issue. And, Dave, do you hit the nail on the head with all the memes going on, of, like,

the the wife seeing $15,000 come out of the bank account? It's like, oh, did you buy something? And then you see the screenshot on the guy's screen of anthropic, you know, credit card payment. It's like, yeah. The people could really run those numbers up if if you're not careful.

Bowen BelmerWell, what I just heard from that exchange was both David and Kyle giving me permission to use the big shiny button for my own personal use cases.

David JohnstonYeah. For personal use cases, I mean, go for it. If if you're staking your your Morpheus, you know, you've got the inference. So like, you know, go crazy. Use your intelligence. Okay. So it's funny that you say that because I genuinely find myself looking at my, like, available daily spend

Bowen Belmersometimes and being like, how do I use it? Like I I wanna use it, you know. I have it every day, like how do I use it? I wanna I wanna be productive. So funny. I don't no token left behind. No token left unspent. Which by the way, in a decentralized marketplace, works because I own the right to use that much inference on any given day. And a right's maybe not the right word because it's permissionless. But the ability, let's say, with no one telling me that I can't.

David JohnstonYeah. That's, that's freedom, brother. Yeah. Don't don't apologize for abundance.

Bowen BelmerYeah. Exactly. Well, okay. So on this topic, well, not not really this topic specifically now. I'm sorry. I'm really jumping around, guys. But going back to you know, we were talking about these localized models like Gemma four twelve b. We do have a celebrity chime in of the week that ties directly into this topic, and it is one of the original YouTuber streamer phenoms, PewDiePie, and his development of Odysseus,

which is a local install, local agent orchestrating tool, basically. And I guess it disbegs the question, win PewDiePie on Morpheus? Because somebody somebody get in touch with PewDiePie is really what this boils down to. Do you That's awesome. Do you think he's aware of things? I mean, maybe not Morpheus specifically. I'd love if he was. But I wonder if you know, he's obviously very deep in the weeds of

owning your own intelligence and why local AI matters. I wonder to what extent he is aware of, you know, the ability of having private AI off of your system as well. Because the reality is, and I think I heard Thomas saying this just as I was jumping back on when I got booted from X, was that there are certain things that you're not gonna be able or want to do on your local device because you're gonna want a heavier heavier load, heavier ability to run more intense hardware.

And so I I wish I'd love to talk to the guy or have someone from the community talk to the guy and get in touch with him and see if he has any plans for anything because this is like the exact type of thing that Morpheus is such a good fit for.

David JohnstonIt looks like he's not an x guy. It's funny. His x profile just basically says, hey. I don't use x, but it still has 600,000 followers.

Bowen BelmerI saw that you zero posts and just an account so that no one else can have the name.

David JohnstonYeah. Yeah. But but we should, we should go dive into his, comment section on on YouTube or whatever his preferred channels are, because I wouldn't be surprised. He he he must know most of the open source tools if he's gone down this rabbit hole. And so he should definitely, you know, be made aware of of Morpheus and and what's available from from the community because it it perp it hooks in perfectly what he's doing. So Yeah. Yeah. And and it's really popped off already. I mean, it's racked up 45,000.

Bowen BelmerIt was 20, like, 23 or 24,000 GitHub Stars in the first forty eight hours, and now it's, like, over 45,000.

David JohnstonNice. Have you had a chance to play with it yet? Would you compare it to, like, a Hermes, or is it is it something else? I have not used it yet myself.

Bowen BelmerBut just reading about it I I guess I yeah. Until I use it, I shouldn't make a specific comment on it. But I I think from what reading though, it seems like slightly different. But I might be wrong. I I I'll have to check it out. I'll have to try it. And and I'd encourage other people to try it as well, because I'd be curious, especially those with a better technical understanding than I have as to how it compares.

David JohnstonCool. But either got a huge audience, so it's it's good, good days for open source and local AI. Yeah. Yeah. Exactly. That's I think that's the big story here. The big story here is

Bowen BelmerPewDiePie and the Pope are promoting decentralized AI with various degrees of knowingness that that they're doing so.

David JohnstonWell, that's, I guess the other thing that came out yesterday or the day before, Brian was saying Church of Latter day Saints has has weighed in, not to be outdone by by the pope and has their own view on AI. I haven't had a chance to read it. Have you have you seen it yet? Well, so the announcement was that yeah. So, apparently,

Bowen BelmerI don't know what to call it. Church watch maybe is what I'm on now. I'm watching to see the various organizations across the world weigh in on AI. And you know, I was making this joke yesterday on a on a call with a couple people, but it is it's it's like a little bit tongue in cheek, but it's also not that I mean, how many times in history, certainly modern history, has the technology been so important and so significant that these really, really large religious institutions feel the need to take a public stance on it? Like, hey, this thing is so big, like we need to take a stance and and so people know kinda where we stand on this thing. I mean, how many inventions have existed where that was the case? A couple, you know, maybe?

I don't know. Unless if I just wasn't plugged into it before because I wasn't paying attention to it. But, like, when when the Internet first came out, did did the churches weigh in on, hey, this is our thoughts on the adoption of the Internet?

David JohnstonI don't remember it. And I was there in the nineties and reading reading a lot. I I think they probably had broader comments on, like, you know, the effects of the Internet or, like, you know, what were people looking at on the Internet or, you know, should their, you know, organization have a website or whatever. But it was it was it was I don't think it was seen in the same way because back then, like, a lot of people thought the Internet was a fad. Yeah. Right? The the the numbers of the .com boom went up in the late nineties, and then it all came crashing down in the February. And I was like, oh, well, that was fun. Yeah. And a lot of people didn't think it was gonna go anywhere from there. Right? You didn't have

even Facebook exist until 2004, broadly until 2006. Right? Pre I iPhone was until 2007, so nobody was really using the Internet on their phones. Like, it took a while for people to realize, oh, like, not only the Internet stick around and a billion people are using it, now there's web two with all the social stuff. Right? And that that sort of launched up a a new re response. You know, web three with crypto. Right? And now you got the AI way. So I I think people are are more savvy now that, like, when these things hit,

they're here for the long term. And whatever precedents get set, like, that's sticking around. So you better have an opinion, and you better be thoughtful about, like, how you're gonna use it because, like, this this stuff is gonna be in the hands of 5,000,000,000 people in a few years. Yeah. Well, I totally agree. It it makes me think of a, kinda two things. The first being

Bowen Belmerthat I completely agree because we have the hindsight of the Internet and how it played out. It yes. It becomes more evidence that they need to be really thoughtful and take a stance on these things that can be as revolutionary as something like AI. The other thing it makes me think though, which is somewhat of a scary thought is we're saying all this with hindsight as to what the Internet became, but at the time of it happening, to your point, a bunch of people did think it was a fad. And I remember even, like, talking to my dad about it. Him saying, oh, yeah. You know, back in especially, like, the eighties into the nineties, the Internet, we just you know, it was like something the government used. I don't know. We didn't really think that much about it. And

I mean, I'm sure that is not true of all households, obviously. But just generally speaking, what what what's kinda scary to think about is, like, that we might still be doing that and probably are in terms of AI. Like, our ability to project the reality of what it actually looks like, it's probably so crazier so much crazier than what we're even thinking because it hasn't happened yet, and we don't have the ability to, like, actually accurately predict it even though we feel like we do. But, yeah, it's no. It's definitely very interesting.

I do wanna say though, so on that point though of the LDS of Latter Day Saints, so the announcement there was that they are holding a sixty minute, like an hour long, like, workshop online this Sunday, the seventh. Where they're that's where they're kinda putting out the official stance. So I will be checking that out. I don't know if I'll watch it live, but I will definitely follow-up and see what it says. Because, yeah, I am interested in

all of these types of organizations, whether it be religious or otherwise, like what the stance is and because ultimately, it it it is like organizations and things that end up being, well, they're thought leaders. Right? Like like, thought leadership does not only happen through religion, but it definitely does happen through religion. It also happens through businesses and governments and all of these other things. So it's really interesting to see, like, how the thought leadership aligns with the technology and and how that kinda trickles down to the everyday user.

But, yeah, I just wanted to mention that because it was another one another one in the belt. And I suspect that the stance will also be inadvertently pro DAI. Whether or not they know it, I suspect that it will be, you know, fall in line.

David JohnstonWell, like we talked about before, like, if if you think about it from a values perspective, you're gonna talk about transparency and, you know, having humans at the heart of it and, you know, wanting all these things. DAI is just how we accomplish that. Right? It's the how to what people know that they want out of this, and it just gives us substance. Right? So, anyway, good stuff. Yeah. Absolutely.

Bowen BelmerOkay. So next topic, just throwing right along here. Worth mentioning, I believe this is another one that there's gonna be deeper, like, more robust update on in not not the short sorry. The near distance to now and not too long of a time from now. But the Manifest and Morpheus integration that's taking place, David, are you able to give any type of update on that? Sure.

David JohnstonSo I'll just I'll tease it a little bit and say I'm I'm ready to spin up the first 42 agents. And this has been in the works for a while, partnership with Manifest and the Morpheus open source developers. We're basically you'll be able to show up to a website, Click a button to log in with Google or whatever your favorite social platform is that creates you an account in the background with privy, you know, so

normal web to social login. Great. Like I didn't have to download anything. I'd have to do anything. Like, you just logged me in. Okay. Now I've got an account. How do you pay for it? Don't have to touch crypto. Don't have to figure out wallets. You could just swipe a credit card on Stripe. Now you've got your agent paid for and what it's doing is it's launching a container, a Docker container on manifest and then the user has access to that agent on that website. So think about that compared to how much friction there is today

to load Hermes or or OpenClaw. Right? You have to download something, you have to load it up on the terminal, you have to figure it out, you have to set up the inference. My goal with this is effectively it removes all of that friction from the user experience so they can just show up, push a button on a website, and instantly they've got their agent. And they can just start talking to it. Right? And tell it what they want and what they wanted to do. And if that's what they want, a web based experience, great. They can just stay using that.

If they want, there's a skill in in the Morpheus skill that lets them export and download all of that to their own computer. So if they set up their agent and they get it all just like they want it, right, and they're they're happy with with using it and they wanna break it onto their local machine, great. Free, no cost, export, download your agent to your own machine. And then you can stop paying for the web version or keep using the web version. Up to you. Right? But this is this is something I've been looking forward to for months and has sort of been putting together all of the pieces,

but in a way where you still really own the agent. One of the big advantages of Privy is it automatically creates a web three wallet for all of the users in the background. So if they along with Google, they've got an address, and if they want to go web three or or download and export all their stuff, they have the private key. And I'm I'm setting up a website for this install openclaw.xyz, and so you can go out on there right now and kinda check it out. Hit me up. I'll I'll post a a

thread on x, and I've already got more than a dozen people that have volunteered to test it out. Like I said, I've I've lined up the tokens for the first 42 toker 42 testers. And so based on people that respond and and get onboarded, I'll I'll get them to test it out. But, you know, for for me, this is a big unlock because I think there's a 100 times as many people that would use an agent if it was on a website as compared to having to use a terminal.

Right? So OpenClaw was a big moment. It made it a lot easier, and you could connect it to a messenger and stuff like that. I just wanna push that to an even bigger audience. And and my my plan is to immediately expand that to to Hermes as well. So, yeah, more details to come. It's kind of thing where I'll probably put together like a whole video showing people the how to onboard. But right now, the call for this week is is testers. If you wanna try it out, hit me up on x. My DMs are open, and I'll get you on the shortlist. But I should be launching here the test version this weekend or early next week, so the next few days.

Bowen BelmerI am so gigabullish on anything that Trojan horses people into using crypto without having to know that they're using crypto. That is just saying. That is fantastic. No. This is gonna be awesome, David. I'm so excited for this and for everyone to give it a try, and for future announcements of, you know, tie ins to what you guys are doing. Because it's it's really stuff like this that drives mass adoption of the technology, you know. Not everyone wants to sit and navigate the terminal to set up their own agent. And in fact, most people don't. And so having the ability to

deploy and just start using is is amazing.

David JohnstonAwesome. Good stuff, man. Well, I appreciate, appreciate the the shout out because it's gonna be a lot of fun to, to kick it off. Absolutely.

Bowen BelmerOkay. Rolling along and to the headline topic of this week, DAI Summer has arrived. So well, all of your bags are nothing but red and just bleeding you dry in terms of value, there is a handful of projects that are ignoring the rest of that, from a token price perspective and a market cap perspective and continue to grow. And essentially,

all of them are either literally DAI or adjacent to DAI, which has been really interesting to see. I'd be curious if anyone else in stage has any stance on why this is happening, or is it just, you know, is it just the case that the productive use cases of the tech and of these different projects are kinda hitting the market now, and that just is what it is. And if that is the case, then what the heck does this look like during a bull market?

Because so many of these projects, you know, including Morpheus, which was done a four x just this year alone. Not to mention things like, of course, Venice, things like I'm trying to think of which other ones I wanna give a shout out to. Near.

David JohnstonYeah. Near. Yeah.

Bowen BelmerIt's it's crazy to see. I mean, do you have any thoughts on this, David?

David JohnstonYeah. I mean, I think we should call this DAAI summer. Right? Like, people have been saying for a long time, like, when are real utility projects going to decouple from just following the price of Bitcoin and Ethereum? Right? Because for years, like, people said, oh, well, there isn't really an industry because all these projects just go up and down based on the price of of ETH and Bitcoin. Well, the last few months has been a a hell of a counterpoint to that, right, where you have the price of Bitcoin and Ethereum going down in a big way, but you have the price of Morpheus,

Venice, Near sort of all taking off in a big way is is a point that, like, we we might actually be there. Right? Real usage, real utility Trump's sort of macro trends. Like, because for these projects, they're burning tokens. They're buying tokens off the market. Right? They're they're shrinking the supply. And so those tokenomics, I think, are are being proven out. But, yeah, it's really, really cool to see because it's sort of been a long time coming where, like, so many people are, like, discouraged

or, like, you know, despondent about the bear market. I saw, you know, Charles Hoskinson said, he's taking a break from from Cartago. And he's like, I don't know what you guys want me to do. It's it's down, and I can't do anything. Right? And so there's so many of these projects that are, you know, going through a bear a bear cycle, and here, DAI has really hit its stride. So I think that's that's so cool to see. So, you know, jump jump on the bandwagon. Like, let's build real utility and make scarcity you know, make make more scarce again. And, you know, I think that that's leading this trend.

Bowen BelmerWell, I, for one, did know that Charles was off break. But regardless, we'll just slide past that. Yes. Could not agree more. And and I think that this this is exactly that decoupling. It is decoupling of productive assets that have utility that people are really using every single day. I mean, I literally use more every single day, period. Like, I just do. And this is the first, DeFi adjacent thing that I can honestly say that's true for. Now, of course, there's other,

like, DeFi protocols and stuff that you could say, okay, you use this every day because if you're taking out, like, a, you know, leverage position through Abe or you're giving yourself a loan against an asset, whatever, like, technically, it's still outstanding. I'm talking about going on your computer and actively doing something that day, every day, that's touching the project. And and I can legitimately say that that is true for me now of Morpheus, as it's true of for tons of people and other projects as well, like of course Venice and

and Nier and and some of the others. So really, really cool to see. I will say though, and this kind of links back to the what we were saying earlier, David, about the boom and bust cycle of the Internet when it first came out and how people kinda thought it was a fad. I have seen some naysayers that have been calling x four zero two a fad because agent on chain payments, like, had this huge blip basically in April into May and now have dropped off considerably. I mean, I suspect it's because people were, like, rearing to try it out,

and now it just has not yet found its exact product market fit. But have you been seeing that as well?

David JohnstonSure. Yeah. I've seen the the stats go down for for x four zero two, but I think probably what that was was a lot of people were sort of gaming it early on for main points, Internet points. Like, every project, you know, made some big announcement about you know, Solana's on four zero four zero two, and and Base is leading four zero two. And then, like, everybody else piled on and made an announcement and sort of got a boost out of being 402 adjacent because it was it was coded for, like, we're doing an AI thing.

But I think they were probably propping up a lot of that that volume. Like, I don't know how much of it was, like, legit. And so and so there's a trough now. But I think what's gonna happen is it's gonna build back up, not based on hype, but based on real usage. And, you know, you see the same thing with, you know, the early testing of Morpheus. Like, a lot of people flowed in to test out the API when it was free. Right? And you had this this huge testing spike. Right? And then the pricing was introduced and, like, okay. Well, here's the real cost. Anybody that was just, you know, throwing volume on there went away,

and then it started building back up based on real usage. So I think that's a healthy process. I don't think there's anything wrong with that. You know, people can, you know, dunk on, you know, x four zero two, but I think it's a it's a real important piece of the tech stack. So I don't think it's going away. I think it's just mean it hasn't found profitable market fit yet, but, you know, I I think it will over time. I mean, I've baked it into the Morpheus skill. If your agent wants to send money, it can use x four zero two. I use it every day.

My agents that are checking price feeds on, like, Queing Gecko, they're using x four zero two. Yeah. And so I'm driving a little bit of that demand. And I think with every agent that wants that kind of like, I didn't wanna pay for another API. And so I pay 1p a day on x four zero two, and I get all that pricing data that I need. Right? But you need a lot of that activity to really add up. Yeah. No. And I totally agree that it will. And it's one of those things that it it'll take

Bowen Belmerthe first handful of agent or agent tools that gain really, really mass adoption and people are transferring money, and it's like, well, how are they gonna do it? They're gonna do it via x four zero two. So it'll it'll happen in due time. I I very much do think that as well. That was all that I had in terms of scheduled topics for this week. So I'm I'm more than happy to open it up and see if there's anyone in the audience who has anything they'd like to add to the conversation or

Thomas, Alan, anything else from you guys that you wanted to make sure we hit on this week?

Alan (Nomadic Rogues)No. Nothing else nothing else from my Yeah.

Bowen BelmerI think we've we've covered the the important topics. Cool. Awesome. Yeah. Alright. I mean, with that, the only other thing I'll add and I've tried to say it every week, but if you're listening to this, first of all, thank you for being here. And second, go out and try stuff. There is now so much stuff to try, whether you are a deeply experienced software engineer who listens to me talk and wants to like jump off a bridge because they're like, oh my gosh, this guy's asking really dumb questions. Or if you're someone like me who might be asking really dumb questions and is like, hey, I want a one click deployment into using my own sovereign agent,

there is a route for you in the DAI space, in the Morpheus ecosystem to get involved. So if you are one of those people or somewhere in between that spectrum, definitely get involved in the social channels, whether that's here on x, whether that's in the Discord. Telegram's kind of a cesspool, I'm not gonna lie to you, so I would recommend the Discord personally. But we do have a Telegram and there I'm just kidding. There are people who are active there. And, yeah, get get involved. And if you have questions for the community, everyone

who I know that is actively working on Morpheus is more than happy to, you know, share the wealth of knowledge and and help get more people involved to express interest in doing so. So, yeah, as always, you know, go try stuff.

David JohnstonThanks, guys. Keep it awesome. See you. Thanks, guys.

Open the transcript file

Hosts and Guests

Hosts

Guests

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Video Is Coming! | Morpheus Is About To Get Far More Entertaining

    52 minAudio

    Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via...

    Credit the people on this episode

  2. Local Models, Foreign Affairs | The Case For Private AI Archives

    53 minAudio

    Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via...

    Credit the people on this episode

  3. Let's Talk Node NEO | Step By Step Progress

    51 minAudio

    Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via...

    Credit the people on this episode

  4. Hello From Hermes! | A New Way To Onboard To Morpheus

    43 minAudio

    Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via...

    Credit the people on this episode

  5. What Agent To Agent Connection Actually Looks Like | Inside The Morpheus Agent Stack

    58 minAudio

    Zion, hear me! For Season 2 of the Mere Morpheus podcast I'm going to be uploading the Morpheus builder weekly updates. My reason for doing this is to make it more accessible for those who want an update directly via...

    Credit the people on this episode