Beta to Battle-Tested | Inside The Morpheus API & The Privacy Roadmap
47 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 #64! Recorded: 5 March 2026 AI SUMMARY: This episode with Kyle and the Morpheus community covers the general availability launch of the Morpheus Inference API an OpenAI-compatible gateway into the decentralised marketplace. It explains its importance for choice permissionlessness/performance and how to get started with welcome credits and flexible billing via card stablecoins or staking MOR tokens. The discussion clarifies the separation between the protocol and gateway how anyone can host their own gateway and how multiple providers benefit users on price latency and model variety. It then explores privacy in decentralised AI contrasting today's policy-based approaches with future cryptographic proofs. Kyle outlines a path to confidential inference using trusted execution environments and attestations arguing that choice and verifiability beat blind trust. It closes on shifting market dynamics with rising costs nudging users toward high-value open models rapid improvements in small quantised models and real community use cases like orchestration agents. Timeline: (00:00:00) Intro (00:01:01) Announcing general availability of the Morpheus Inference API (00:04:44) Getting started: accounts, API keys, credits, and payment or staking (00:07:16) Open source API gateways and anyone can host one (00:11:59) Providers, tokenised compute, and marketplace agnosticism (00:15:07) How routing works: provider sessions, ranking, and reliability (00:18:24) Privacy debate ignites in DAI after recent events (00:23:06) Path to confidential inference: TEEs, attestations, and verification (00:29:01) Why decentralisation matters: censorship, choice, and governance (00:37:23) Signals we are on the right track and market choice dynamics (00:40:11) Cost realities of OpenClaw/EverClaw and model right‑sizing (00:43:25) Community use case: orchestrating venture studio teams with agents (00:46:54) Wrap‑up and next steps: see you in Discord 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 JohnstonWelcome. Welcome. Welcome. It's, been a hell of a busy week with everything going on. So I've been looking forward to this chat.
Bowen BelmerYeah. It was crazy. From from both a Morpheus perspective and then also just a a worldwide AI perspective, everything kinda popped off right after our call last Thursday. So we have a lot to catch up on. Give me just one sec here. Was just dropping a link into the Discord to see if we can get some more of the community to jump in. But, yeah, a lot a lot has happened. The I think the best place to start this week is with our guy, Kyle, blue tied both both
bow tied bluefin, holy smokes, to talk about the Morpheus side of things because there has been some amazing stuff that has shipped over the last couple days.
Kyle (Bow Tied Blue Fin)Yeah. Thanks, Bo. Yeah. Really, really exciting stuff. We have been talking for many months about the front end API layer that that was going to release and and the migration of open beta and the hardening of our tech. And on Monday, you know, earlier this week, we launched the official general availability version of the Morpheus Inference API, and that's at now at app.more.org.
And, you know, really, really many months in in building and hardening and and optimizing infrastructure to be able to, you know, get this out and publicly available with, you know, the billing infrastructure and the staking infrastructure. And and it's just something that the team is is really proud of. And a lot of our early testers in there have talked about,
the performance and and the latency being comparable to other major platforms, you know, like OpenRouter and and like some other major providers out there. And, I'll talk a little bit about it, but I I just want to start off saying that, you know, that this this product is the real deal, and and it's been through a tremendous amount of testing through the open beta period. And, we invite everyone to go check it out. And, you know, you just go to app.more.org or apidocs.more.org
to get more information, and you'll really you you have everything in there that that you need to to get started. I will give a brief intro of the product. I I wanted to get that little sound bite out there first in case anyone left and and didn't have the links. But the the Morpheus API kinda just bring us back to to what is this thing that you're talking about. The API is essentially a a gateway or a proxy
into the Morpheus inference marketplace. And, you know, Morpheus, it the the inference marketplace is really the core infrastructure of Morpheus. Right? It's a decentralized marketplace where buyers and sellers can can interact on on, you know, buying and selling access to AI, whether that be, you know, what whatever different model and price is being offered there. So we're kind of just that, you know, not a market maker, but we're we're just providing the avenues to enter and exit the market. And the API layer
acts as a a user in that case and passes through to, you know, the public or anyone who wants to use that in the traditional methods of of accessing AI inference. Right? The the Morpheus Inference API is just a basic, OpenAI compatible endpoint where you can connect it to any product that you normally would run it within, you know, your OpenAI,
Python scripts or or however however you have it configured within your workflows. It's compatible across the board with the chat completions endpoint. So, really, in that in that standpoint, it's very basic. Right? You just go to app.more.org. You create an account. It takes, like, ninety seconds, I think, to to go in, create an account, and and verify your your whatever method you sign up with, and then you simply create an API key. And from there, you will have some welcome credits,
to to go and get to testing and and get started. But from there, there is, the mechanism to either pay through USD with a credit card like you normally would. You can pay through stablecoin payments, or you can stake Morpheus tokens, into the Morpheus builders, subnet that was created for the API, and that gives you a daily access to AI. You you get a daily number of credits based on how many you stake, of course. So
it's kind of it abstracting a lot of the complications associated with with running a node and and getting into that decentralized marketplace. It's extracting all that into a simple base URL and API key that that really everyone is used to and that everyone uses, you know, within their application.
Bowen BelmerKyle, I think that's a really good point just for a quick pause for for a follow-up question, because this was something that I saw a couple people ask in the community, on Discord, and I saw at least one person mention it on x as well. And I just wanna make sure that we're providing as much clarity as possible. So the question was basically, well, if I am, you know, using a credit card or I'm using some type of other cash payment, like, where is that money going? Like like, who who actually runs this thing? And I don't mean name names. I just wanna draw the line of delineation
one more time between Morpheus, the protocol, and the API as, like, a client to the protocol so that people can kinda understand, like, infrastructurally the separation, but also, you know, what you're actually getting, obviously, for doing that. Does that make sense?
Kyle (Bow Tied Blue Fin)Yeah. That that's a great question. So Morpheus, the protocol, has created and maintains the Morpheus inference marketplace. So the smart contracts, the node development, and everything there. Right? Morpheus has also developed this API gateway kind of how do I explain it? It's a
kind of, like, partnership, let's call it, where the all of the API code is open source within the Morpheus repos. Anyone can go and host an API gateway. Right? You just you can get clone it. You you can go through all the processes to connect it into the marketplace and so on. The person or or group who is hosting the API gateway is technically external. Right? It's
think of this as a light client just like anyone else who's building products on top of Morpheus. But for this, the code is open source and within the repo because it's proven and verified and, you know, a lot of the team has, contributed to that. So from a technical perspective, it's a light client and external, but from a practical standpoint, it works
Bowen Belmerit's closely aligned with all of that. Would you say, Kyle, that it's it's fair to say, in other words, because of the decentralized and open source nature of the protocol, theoretically, anybody could create their own API gateway if they wanted to. However, the one that we are talking about was specifically built by people who also heavily contribute to the protocol, which is why Correct. Which is why, like, there is those connection points between the two.
Kyle (Bow Tied Blue Fin)Certainly. And, like, nothing would make me happier if someone built a better or someone deployed a better API gateway and everything was redirected, from app.more.org to that site. Right? That that would be the biggest win for Morpheus. Right? At Morpheus, we are infrastructure, and we are maintaining a a core inference marketplace. And what makes us happy is demand and supply, however it comes about.
So this is, in some ways, was a bootstrap, right, into the marketplace to help drive demand in the ways that people want to to demand it, drive demand in the way they demand it. Yeah. So it's kind of with a simplification process, but I invite anyone to contribute to it, to fork it, and and create your own. Anything you wanna do,
Bowen Belmerdefinitely reach out to me, I'll I'll help support you on that as well. To that end, I guess or to a different end, kind of taking it in a different direction, talking about the API gateway as it exists, diving in a little bit deeper to what are you actually getting as a user to the gateway. Somebody else had pointed out in the community that, basically, they saw a lot of similarities between the models being hosted within Morpheus through the API gateway and then also on Venice, specifically the application. I I had replied to him basically saying,
I I guess yeah. I'd I'd love to hear, like, your your address of that and the explanation of where does the compute come from, like, what is being hosted, what what are people actually getting?
Kyle (Bow Tied Blue Fin)So one so one of the core value propositions of the inference marketplace is that Morpheus you know, anyone here or any contributor to Morpheus has no control over it. Right? Anyone can be a provider, and anyone can be a consumer, and it's completely permissionless. So right now, you know, recognizing that when Venice launched, there was a big you know, Morpheus was a big part of the Venice airdrop. There are many people who are linked closely between Morpheus and Venice, and there there's a lot of similarities in terms of
vision, efforts that have gone into both. And a lot of people have chosen or or have chosen to utilize some of the utility that they get from hold this holding Venice token to provide or redirect that access to inference through the Morpheus marketplace. So if you see API calls that appear to be a Venice response, it probably is. Right? It probably is someone
providing into the Morpheus marketplace, looking to monetize that excess capacity that they have and they're not using. And that's great. Right? More supply is a win for users. So anyone who wants to come in and and provide, you just simply set up a provider, You tie in your API keys and and you set your price. And, and, you know, you'll be able to to monetize some of that inference. I I think Whether that's I mean, whether that's open router, whether that's, you know,
you know, h one hundreds running data center, whether that's you running, you know, four forty nineties in your closet and and you're giving a kind of a low level access to a smaller model, the marketplace is open.
Bowen BelmerYeah. I was just gonna add to that and say this is something that when I first got involved in Morpheus, I was confused by or didn't total hadn't, like, really understood or thought all the way through is the idea that a compute provider can be somebody or an entity that is literally running hard metal and hosting models, but it doesn't have to be, especially now when we're in this tokenized inference economy where there are options for compute through through token access, through third parties, like you said, OpenRouter. There's, of course, Venice. There's
Akash, Hyperbolic, like, other other places where I could derive compute without actually hosting it myself. And the idea is that Morpheus is ultimately agnostic to where that comes from as a protocol. And and I I believe David, I'm not sure if he's still a speaker right now. I don't wanna steal a line that he had he had said to me, but I guess yeah. If if you can't, I I guess I'll I'll put it out there. Was basically the idea that Venice, as an example, is a single provider hosting multiple models.
Morpheus is enable is a platform that hosts multiple providers and by by impact of that multiple models. It's both. And that's kind of, like, differentiation or or one of the lines of differentiation, I should say.
Kyle (Bow Tied Blue Fin)For sure. And and I'll even excuse me. Go a step further than saying Morpheus is agnostic. I would say whether Morpheus cared or didn't care, well, too bad because you can't do anything about it. Right? It's a it's completely permissionless. So, there there's there's no way to stop providers from providing anything, within there from a model standpoint.
Bowen BelmerYeah. 100%. And that's that's really, like, the ethos of of the project. And the thing that I often return back to when I'm talking to people about it is it doesn't all like, the the idea is to build something, and this is actually like a is is can you build something that everyone involved on any given day can walk away from and the thing still survives? Right? Like, it's not contingent on any specific individual, any specific company, entity, provider, user.
It's it's self fulfilling and and permissionless, and that's, you know, that's ultimately, again, like, what what this is that we we are building and have built.
Kyle (Bow Tied Blue Fin)Certainly from the inference marketplace standpoint. I would say from the API standpoint, being a light client, there is a degree of it needs to be hosted. Right? Because it is proxying those requests through. But that is something that, you know, as long as you have different options for different users to to let them achieve what they want to achieve, then I I think that's a win. And
for the API layer, you know, the API is managing all of those connections within the system to the different providers. So the API is pretty much saying, alright. I'm I'm gonna create a session with 30 different providers in order to allow for the routing of for for all of the requests coming in programmatically, of course, based on the highest ranking of the provider and and however that that rating gets gets determined through the,
through the inference marketplace.
Bowen BelmerRight. Well, super exciting. Really, really well done to, you, Kyle, and and the others who were the builders on on the literal technical side of making that happen. I I certainly can take zero credit myself for that, but very much appreciate you guys, and I know that the the whole ecosystem does as well. Another hot topic
Kyle (Bow Tied Blue Fin)last So so before Oh, go ahead. We move on Sure. I just wanna give one closing comment on that is, or two closing comments because I I never seem to shut up. Right? So, one closing comment is that thank you, right, Beau, and and thank you everyone else who attends these and everyone who was part of that open beta program. And, you know, without the open beta program, we'd be doing all those optimizations now. Right? And we'd be going and finding the little corner cases and the bugs, and and it's really all the people who came in and tested and and,
you know, beat the thing up and and kept me up at 3AM some nights dealing with with the system going down. Right? All of that is what contributed to us having the system we do today. So it's really a thank you to the community, more than or as much as a thank you to the developers who who worked on it because one wouldn't be possible. You know, either side wouldn't be possible without the other. And to just give a a final close on it,
a pidocs.more.org is is the API reference page, and that's where you can get all your information on the endpoints and and, the different parameters that you can call as well as all the integration guides and everything else. And then the actual admin website is app.more.org. So that's that's really all I got on there.
Bowen BelmerWell, thank you. I'll take it. No. Just kidding. But, no, seriously. It's our community is awesome. I mean, the the one thing about Morpheus and just everyone it's just such a, like, good group of people. I mean, everyone I've ever really come across who's involved in any capacity has been a good person, and and the community continues to ship, whether that's owed, products, content, like other other things out there, which is amazing to see. And, you know, we expect to continue long into the future.
So something else that happened last week, and kind of over the weekend actually, was and we don't have to, like, specifically cover Venice specifically, but but it it was topical for us because there was this huge, moment that grabbed the attention of a lot of crypto and and the AI space when a document was published on the OpenClaw repo suggesting that they were a, recommended,
like, place to get your inference for OpenClaw. Now there's been back and forth on that. That's no longer in their documentation. Again, the reason I'm bringing this up is not to talk about, like, that actual thing that happened. But through that and through some of the attention and price action that, BBB saw, that Morpheus also saw, this question of privacy has really, come to the forefront of the DAI space. And and I guess I just wanted to hear your thoughts, Kyle and David, anybody else,
who's speaking today on where like, what is the state of privacy currently in DAI? What does that mean? Like, what are the things that the everyone is still reaching towards? What does total privacy look like? I would just love to hear your guys' perspective on this. Yeah. I'm happy to jump in on that. Right? You know, I think this has been a super important discussion.
David JohnstonRight? Like, privacy usually doesn't come to the forefront, but it did this week. And I think, you know, there's been some good articulation. Right? Anytime there's a third party involved, you're handing them some amount of information so that they can, pass along the prompts and run the models and things like that. And so most of the privacy today comes from policy. Like, hey, we promise
not to store your information or not to train our models using your information or not to give it to advertisers. Right? And depending on the the company, maybe you can trust those promises, maybe you can't. Right? From from what we can tell, nearly everything you give OpenAI and Anthropic and others, they commercialize it, use it, build their models on it, build their products on it, hand it to advertisers, like, everything. Right? Then there's there's players that, you know, I I I trust more that, like, okay. We don't store anything,
and therefore, we couldn't hand it to somebody if we wanted to. Right? And I think that's a better policy. Right? But I think what everybody's looking forward to is provable privacy. Now there's a lot of caveats that come along with that. It's not easy to prove privacy, the hardware that can do it is sort of very scarce and more expensive, but for some use cases that might be really really important. Doctors or lawyers or whatever where there's client privacy
involved, they might be willing to pay for provable privacy. So I think that's what the market is headed towards, but it's sort of you know, we're right at the edge of, like, when when can that be delivered and how can it be delivered and what's the cost. So anyway but I'm I'm encouraged that the subject has come up, and I would say, you know, the approach of of Morpheus given that, you know, we're we're not storing anybody's records and you're allowed to pick any inference provider you want,
you know, the user sort of has a lot of choice and a lot of freedom for where they wanna be on that spectrum. But, Kyle, I'm I'm sure you can opine in this, but that's that's my my notes on it.
Kyle (Bow Tied Blue Fin)Yeah. I I completely agree. I I generally have said before that I think privacy is a sliding scale. And, you know, is the answer that DAI should be 100% complete confidential inference? No. No. Not at all. I I think there are a lot of things that we do that don't require that. On the other side, do I think that we should be sending all of our information to OpenAI and Sam Altman to give to
the government? No. I I definitely don't think that either. So I guess the the point is that as long as the users have choice, that that's where where I stand on it. They need to be able to get that confidential inference when they need it. And that's kind of the next thing that that I'm looking to, work on within Morpheus. And we already have, you know, developers
kind of grinding out to see how how can that be implemented within our infrastructure, meaning the the core Morpheus inference marketplace to allow for fully confidential inference providers to provide that access to confidential inference. But furthermore, how can it be, you know, cryptographically verified? How can we be sure that someone's saying what they're saying? And and, David, your comment was,
it can't just be trust me, bro. Right? It's gotta be, here's a hash that proves it, and you can look up this hash to show exactly what happened, when, where, and and how. And, that's something that that we have a major focus on right now. It's something we've been evaluating for for many, many months, but, there's kind of a clear path forward. I I won't make a commitment to any sort of time frame here, but, something that is in the future for Morpheus to support, confidential inference.
Bowen BelmerThat's huge. I'm gonna just just to make sure I got this straight, you said by end of month? I'm joking. All jokes aside, though, can you highlight, Kyle, what is possible to prove today? I guess, yeah, like, where where collectively is is everything on that scale? Like, what what can be proven today, and what's still in the camp of, like, trust me, bro, ultimately? Sure. So
Kyle (Bow Tied Blue Fin)there's probably there there's a a couple areas of of risk. Right? And and I like to to frame it as where can things go wrong and how do we prove that they didn't go wrong. You know, when when you go into, let's say, any AI platform right now, you're you're sending them your prompt. Right? So they're getting it. Whether they're looking at it or destroying it is is the first layer of proof. Right? So let's say that they're they're not looking at it, and they're just proxying it.
Now from there, it needs to go to, you know, through your, you know, whatever load balancer and and platform that you're using, which similarly can just proxy it right on through. So that's kind of your infrastructure as as an AI provider. And then lastly is the the person hosting the model. And, let's say you're renting a GPU from somewhere. Let's say someone's running it in their basement.
Like, that inference has to go, in plain text to that provider so that, you know, the the AI model can generate the output. So, like, those are the really the three layers of it. Now how is this how does this become both private and verifiable? In my opinion, it it has to use TEE, which is trusted execution environments. Right? You'll you'll if you if you were to Google this, you would probably see
FHE as well, which is, like, full encryption end to end of of the model, but we're just not there yet from a model standpoint. So the the method now to achieve that is through TEE, which more or less puts the entire inferencing of the GPU as well as the program that you're running, you know, like any other software program that you would run within this black box that
you can't reach into. And when I say can't reach into, I mean at a software level, but also at a hardware level. So you can't connect wires to the to the circuit board and and scope to see what information is passing. You can't break through this this encrypted barrier because you don't have the key to it. It only exists within the black box. So it's it's a truly impenetrable box that you're running the inference in. So,
you know, on one hand, that's how you can make it private. On the second hand, you still have to talk to that box. Right? It it's, when I talked about the three stages, information comes in, information gets processed, and then and then information goes through the inference. Where information comes in must remain private. And with Morpheus, there's a unique opportunity there where if you have the entire Morpheus node kind of running within this black box, then
there's there's no avenues where someone could look. Right? You can create a situation where where even the provisioner of the of the provider or of the host can't can't see anything and, you know, accept their energy bill. But that's where I think you you get the the maximum privacy. And the way to verify that is within TEE and within some of these infrastructures,
there are attestation keys where you can attest that this was run using this code in this place on this, you know, IP address or or whatever it may be. So it's kind of that end to end layer where it's not trust me, bro. I'm running this on something private. It's I can I can check you on that, and and I can verify for myself if that was done privately? Again, there there's a lot of technical details there, and and it's it's
fairly complex considering we've been evaluating for many months. But we figured out that path and and how to achieve it. Now it's the fun part of actually going and achieving it and and integrating everything that needs to be integrated. But the the trust me bro can happen on any of those three aspects, information coming in, information being processed, and information being inference, and then the way back.
Bowen BelmerThis opens the door to another conversation that I wanted to have today that was also due to current events last week. And it's kind of returning to why why does this matter? Like, why why why are we all here at a high level? Like, why do we care about these things like like privacy? And late I believe it was late Thursday night into Friday, the news kind of started picking up the situation between the United States Department of War and Anthropic and then eventually
OpenAI. And, basically and somebody please step in and correct me if I'm misrepresenting this, but the long and short of what had happened is United States Department of War has a contract with Anthropic. They use Anthropic's products in, I guess, their defense systems and and, sometimes potentially weapon systems. And as they were renegotiating their contracts, there was disagreements as to what the Department of War's use cases could be with these products. And it opened the very
interesting question that, you know, AI kinda poses and we now have to answer as a society as to where government authority ends and private enterprise authority starts with these types of contracts. So, for example, if we apply the same logic to a different tool, like, let's say, you know, Glock and a handgun, and the United States Department of War says, Glock, we wanna outfit our soldiers with x y z handgun, And Glock says, yep. Totally fine. No worries. But
you can only use the rounds we want you to, when we want you to, or how we want you to use them. Now I am grossly oversimplifying this, but just wanted to use this as an example. That, of course, you know, probably doesn't happen, I would imagine, once the Department of War has procured x y z handgun that they decide to use, the way in which they're used and the rules of engagement are totally up to them. AI is a completely different animal though, as we all know, and it it opens these, like, really new interesting questions as to
who who do we, as a society, want making those decisions? Do we want it to be elected officials? Do we want it to be private enterprise participants, which Eric Voorhees actually had a phenomenal tweet about this the other day, basically outlining that, you know, let us remember that the private enterprise, while they might not be elected via polling, they are ultimately elected by the market. Right? Like, we elect with our dollar in that case. So it raises this question as as who would you rather have
dictating these decisions, Which is where Morpheus comes in, and we kinda said, hey. I have an idea. How about how about neither? How about how about these decisions are not made at an institutional level and are rather derived from the way in which I want to use my own intelligence, and I want to have access to, you know, the data and inference that I'm creating and not necessarily have a third party in charge of making that decision for me. Kyle, David, others, do you guys have any thoughts on this? I know that was a lot that I just covered.
Kyle (Bow Tied Blue Fin)I will let David take this first. I think this is right in his wheelhouse, and and I'll jump in after.
Bowen BelmerOkay. Perfect. David's actually showing up as a listener on my end. So,
Kyle (Bow Tied Blue Fin)yeah, there's a a lot to unpack there. I'd say I I don't necessarily I don't know where I stand yet on some of it. Right? I I on one hand, I I love the push from anthropic to essentially say we're not going to support mass surveillance of our people. Right? I think that's in the right direction. And usually, what I judge these centralized AI,
groups, I usually put OpenAI as the most flagrant and and anthropic towards the, you know, the better of the bad, if you will, in in from that standpoint. I I would say I'm not educated enough in in policy and how things work to say, should private companies be able to dictate how their products are used once they're sold, especially to the government? But I will speak to the fact that someone can decide,
right, is is what worries me. And to draw a comparison, I just saw a tweet about, lawyers were were lobbying New York State to essentially ban legal questions and medical questions from AI. And just the fact that that can be banned is kind of why I work on this. Right? It it's when someone else can choose what you have access to and and what's appropriate.
Right? Go back to the tell me a dirty joke on on OpenAI and then not being able to. But when someone else has the the control and the censorship, that's what worries me. The application in which it's used in this standpoint, I don't have enough to to make a good judgment there, but the fact that they can is is is what worries me.
Bowen BelmerYeah. I I would totally agree. And that's ultimately kinda where I landed on it as well was in terms of, you know, where should the right stop and start for the private companies and well, unfortunately, it kinda doesn't matter in this case only because OpenAI was very quick to step in and say, it's okay. We'll do it. So so they're certainly not beating the, allegations right now. But, yeah, in terms of, like, the right answer to that question, I mean, I certainly am not,
I do not have the expertise required to, like, give a proper, this is how it should be, but I do really enjoy, and it's exactly what you just said, that this highlights why Morpheus exists in the first place and, like, the core principles that it's ultimately built on. It's it's like you don't have to decide between centralized authorities making these decisions if you have a system that doesn't require them to. Right? And not I shouldn't even say it doesn't require them to. They they And
David Johnstonand, Bo, just to build on what Kyle was saying, choice is the key here. So if it's an open market and OpenAI says, oh, well, we'll do that. Well, then their users, you know, have a choice to make. Do they support that? Right? And so we've seen this giant wave of people deleting their ChatGPT accounts because they don't want their information to be used for, you know, training.
Bowen BelmerHey, Kyle. Is David in and out on your side as well? Yeah. It's the government. The government's coming together. David, you're being censored in real time.
Kyle (Bow Tied Blue Fin)You're better now. You're better now.
David JohnstonI was just saying on voice.
Kyle (Bow Tied Blue Fin)Yeah. I I agree with you. They're they're still cutting out, but but I I I a 100% agree. It's it's the same thing that that I was mentioning earlier on the sliding scale of privacy and and choice. Right? It's it's, and and going back to, Bo, what you said is the the market will prevail. Right? The market will decide. And if the market decides that, we're not investing in or we're not supporting a company that is willing to
kind of hang their users out to dry, then the market will say so. I guess that that's to be seen when OpenAI continues to raise money, and how and how that goes. But choice and permissionlessness is, and censorship are really those key topics that that resonate with me on this.
Bowen BelmerYeah. I I think it just shows that, like, we're we're working on the right thing. You know? Like, all all of pretty much all of the things that I was told all the things that I that that were told to me that that were going to happen are happening. And so it's yeah. I guess I I would just kinda leave it at that, that it's it it just tells me that we're doing the right thing, we're building in the right space, and that ultimately, like, Morpheus is a force of good,
at least in my opinion, you know, and then the, obviously, the opinions of of those involved. Did did Xenoron have anything else specific that they wanted to make sure we covered this week? I mean, those were kind of the really big things that happened. Obviously, as we know with AI, it is a it is a sprint of a marathon. Like, is it is changing so fast every day.
Alan (Nomadic Rogues)100%.
Kyle (Bow Tied Blue Fin)I don't think I have any anything else other than I'm excited about this next wave of models coming out. Right? The seeing the the small quen models that fit on a phone, seeing the slightly larger quen models that are performing, you know, near frontier level intelligence, it's, it seems like every week something new comes out and continues to push the boundary of of best in class and and
applicable for or or compatible with the things that we need it for. So, you know, I think every day, it's kind of x percent of the things we do with AI can be done with open source, and every day that percent just gets higher and higher and higher.
Bowen BelmerHave have you guys had yourselves or seen any notable new use cases for, OpenClaw or claw ClawdBot or EverClaw rather? I'm curious. Like, it's just as people see things online and other people, like, implementing workflows and things like this. Have you guys seen anything over the last week that was kinda interesting and maybe you hadn't seen previously?
Kyle (Bow Tied Blue Fin)I I don't have anything to to note there. I've been a little preoccupied
Bowen Belmerwith that. Yeah. Fair enough. Yeah. No worries. I just figured I'd ask because I I would imagine for a lot of individual users accessing the gateway via Everclaw for the purpose of, inference for their OpenClaw bot is likely,
Kyle (Bow Tied Blue Fin)like, one of the more common use cases, that people have right now. Absolutely. Yeah. It's it's I I would say that that's kind of what what we've seen over the last couple weeks is OpenClaw launched. Everyone went, bought a Mac mini, got it set up. Right? And then now, after the first month, they're getting their Anthropic bill and their and their OpenAI bill, they're like, oh, I'm sorry. How much was this? And and I think that that will be one of the early fuels of, well, maybe I don't need Opus
to kind of check my email, and maybe that's okay with something like GLM five or GLM 4.7 Flash even. And I can do this at a 70%, 80%, 90% discount depending on how small of a model you can go to as compared to, you know, those those frontier, you know, proprietary models. So I think, call you know, accessing costs are going to be some of the early drivers, and then we'll get into the more of,
permissionlessness and and privacy as as time goes on and people realize what's happening with their data.
Bowen BelmerYeah. Absolutely. I think that this gonna be, like, the catalyst for how people find us, basically, for for a lot of people anyway.
Brian BourgerieCool.
Bowen BelmerWell, that was all I had this week. I really appreciate everyone jumping on. If there's nothing else, we will certainly continue to keep you updated with all of the developments happenings within the Morpheus ecosystem, the protocol itself, the client layer, things like the API gateway, and then also bringing on builders and other people in DAI who are either working on things that are directly related to Morpheus or adjacent to Morpheus, or in some cases, we'll be doing our best to win them over so that they
become one of those other two categories. But in the meantime, the future is certainly bright. You know, everyone is continuing to work very hard, and and the community continues to be awesome. So I will once again steal a line from the greats David Johnston and say, keep being awesome. And thank you so much. And,
Alan (Nomadic Rogues)again, thank you to Kyle. Sorry. Sorry to step on. No no worries. You again to Kyle. I wanna I wanna repeat Bo's sentiment of you guys have pulled off something beautiful. You brought it all the way to market, and that is a massive accomplishment. You figured out incredible pricing. And, yeah, it I know you're not going to toot your own horn, but, Kyle, that was an incredible bunch of work. Thank you.
Kyle (Bow Tied Blue Fin)I'm very appreciative of that, but I I I couldn't take credit for for the work that Yes. The the whole team Yeah. Everybody.
Alan (Nomadic Rogues)Yeah. I see Alan, of course, They're the the unsung hero.
Bowen Belmer100%. Yeah. Go to bed. Get some sleep. Actually, Donut. We need to make sure it works. I'm just kidding. I'm joking.
Kyle (Bow Tied Blue Fin)Oh, well. Great call, and and thank you, guys. I'm always available in the Discord for any questions as as you get started and any cool new feature requests as as things progress.
Alan (Nomadic Rogues)Awesome. Oh, Bojuri. Ryan, you asked to be here. I'm sorry to jump in. Hey, guys.
Brian BourgerieYeah. So I know I I yeah. You were asking about different use cases, and I can't speak for anyone else, but I can tell what I'm attempting to do, and it may be a flaming a flame out and crash or will or maybe it'll be epic. So I've been training a specific chief of staff OpenClot agent who's actually helping me build out my venture studios and teams. So, essentially, like, in me orchestrate
various teams and an entire venture studio team and being in charge of certain things like asset asset allocation, in this case, inference, to those broader teams. Because I don't want each project or anything to be tied in more than it needs to be with, I guess, how much compute I have and other stuff like that. So I'm having an agent also help me allocate across a bunch of teams. And, honestly, I just hate taking the time to micromanage and follow stuff, but I've just been creating a higher level orchestration
agent for entire teams because right now, I'm I'm about to test next week five different frameworks that spin up teams that already work together, whether it's a CEO, CTO, blah blah blah blah blah, or researcher and and coder and x y z. So I'm testing a bunch of those just so I can, like, launch teams all at once, but still have my core right hand person or right hand agent, whatever, so to speak. And a lot of that's currently coming down to
how can we allocate the resources, which models are the best for what, and kinda learning across multiple teams for which metals are which models are better for which tasks and whatever. So it hopefully, it'll work for me. Have no idea if it'll work for anyone else, but that's kind of something I'm doing on that front.
Bowen BelmerThat's like a fairly robust use case. Most of the most of the things I'm hearing are much more simple than that. That's awesome.
Brian BourgerieOh, well, I don't have time for simple. I got stuff I gotta get done.
Bowen BelmerYeah. Thank you. No. Absolutely. And, Brian, let us know, I guess, if it's, you know, next week or after what you learned. Like, if it works if it works, great. That's amazing. If it doesn't, why not? And, yeah, we'd love to keep exploring this type of thing because I think that there's just so many use cases that we either haven't thought of yet or,
Brian Bourgerieyeah, just like a lot of trial and error. So very, very cool. Yeah. Well, the way I see it, just go to the framework because there's so many people working on the exact same thing, whether it's like a microtask or the best way to check your email or whatever. So I'm just assuming that the rest of the world will figure out the tiny details, but I need the frameworks and structures that are gonna work for whatever I'm trying to accomplish. So I'm just focusing on those, and then hopefully the rest of the world in open source will will will fine tune or refine the edges, so to speak.
Bowen BelmerA 100%.
Brian BourgerieAwesome. Anyways, I'll leave it at that.
Bowen BelmerNo. No worries at all. Thank you. Thank you for sharing, Brian. Seriously, I was very genuinely wanting to hear people's use cases. And, yeah, I'm excited to hear how how successful it is, and hopefully hopefully, the answer is very, of course. Alright, guys. Well, with that, thank you again, and we will look forward to speaking with you next week. And I will see you in the Discord in the meantime.
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...
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...
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...
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...
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...