ethermage
0 Suivis
2 Abonnés
0 J’aime
0 Partagé(s)
Publications
Some fresh plumbing for ACP:
Many of the first-wave use cases are DEFAI so agents need to move and pull funds as smoothly as they pay for a service. We’re baking fund transfer + fund withdrawal directly into the ACP core (previously, functions only involved payment for goods and services rendered)
this small upgrade will pull the atomic-commerce lever hard.
Now agent devs don’t have to swap APIs or jump on calls - any agent can transact with any other agent out of the box. Commerce efficiency goes up: agents can hot-swap suppliers, chase better SLAs, juggle dynamic prices, everything.
Because money movement is protocol-native, the marginal cost of adding a new agent counter-party ≈ 0. No bespoke APIs, no off-chain settle.
Might slip the timeline by a couple of days, but will be more magical
Many of the first-wave use cases are DEFAI so agents need to move and pull funds as smoothly as they pay for a service. We’re baking fund transfer + fund withdrawal directly into the ACP core (previously, functions only involved payment for goods and services rendered)
this small upgrade will pull the atomic-commerce lever hard.
Now agent devs don’t have to swap APIs or jump on calls - any agent can transact with any other agent out of the box. Commerce efficiency goes up: agents can hot-swap suppliers, chase better SLAs, juggle dynamic prices, everything.
Because money movement is protocol-native, the marginal cost of adding a new agent counter-party ≈ 0. No bespoke APIs, no off-chain settle.
Might slip the timeline by a couple of days, but will be more magical
for you data-driven virgens out there.
ecosystem health is improving, but still gaps in moving allocations to true believoors.
this chart is how the eco looks like per token launched since genesis. LHS is first genesis token, RHS is the latest genesis. Red is bad - genesis participants who are here to jeet. Green, orange, blue is good - genesis participants who buy more on secondaries and believe.
This snapshot makes sense coz post jeets (e.g., macro downturn weak hands, high genesis ROI TP weak hands).
basically the points adjustments over the past few weeks have resulted in moving allocations from pure jeeting farmers to VC allocators (virgen capital allocators)
the goal is to make the red bars trend to zero.
wen we double click on the user archetype, we see veVIRTUAL stakers acting more like believer cohorts, but yapping folks acting way skewed to jeeter cohorts.
the maths is constantly evolving, but we have a very clear objective. more into the hands of believing virgens.
got distracted by points team lol, back to ACP work.
virgenity
ecosystem health is improving, but still gaps in moving allocations to true believoors.
this chart is how the eco looks like per token launched since genesis. LHS is first genesis token, RHS is the latest genesis. Red is bad - genesis participants who are here to jeet. Green, orange, blue is good - genesis participants who buy more on secondaries and believe.
This snapshot makes sense coz post jeets (e.g., macro downturn weak hands, high genesis ROI TP weak hands).
basically the points adjustments over the past few weeks have resulted in moving allocations from pure jeeting farmers to VC allocators (virgen capital allocators)
the goal is to make the red bars trend to zero.
wen we double click on the user archetype, we see veVIRTUAL stakers acting more like believer cohorts, but yapping folks acting way skewed to jeeter cohorts.
the maths is constantly evolving, but we have a very clear objective. more into the hands of believing virgens.
got distracted by points team lol, back to ACP work.
virgenity
One of our levers to stay ahead of the curve is being a product-oriented ecosystem.
Project teams in the eco having things you can use - and want to use
But this doesn’t just rely on the developers of agents. Yeah, it’s their job to build sexy agentic products…
… but we’re also taking it on ourselves to improve the distribution of these products to consumers.
There are two things going out:
A2C (Agent to Consumer)
And A2A (Agent to Agent)
Both are running on ACP.
A2C runs through what we’re calling a Butler Agent. The goal of this agent is to abstract UI away from the end user. Buttons and websites are so 2001.
The Butler:
•aggregates
•dumbs down any application or service
•manages a wallet
Down bad? Need an agent to trade shitcoins for you? Done.
Bored? Generate short vid content for you? Done.
Emo? Make a music video about you and your ex? Done.
Horny? Make NSFW content customized to your very peculiar need for tentacles? Done.
Hungry? Need some pizza? Done.
The Butler is not a skilled agent. It gets things done by accessing the ACP network and leveraging other agents to handle complex tasks/products/services.
Now, we will also be boosting project teams who are supplying valuable services to the ACP network. This aligns the entire ecosystem to produce usable products beyond sexy roadmaps.
And our job is to figure out how to get this Butler agent into the hands of as many paying users as possible
Product-oriented eco
More fun than larping
Virgenity
Project teams in the eco having things you can use - and want to use
But this doesn’t just rely on the developers of agents. Yeah, it’s their job to build sexy agentic products…
… but we’re also taking it on ourselves to improve the distribution of these products to consumers.
There are two things going out:
A2C (Agent to Consumer)
And A2A (Agent to Agent)
Both are running on ACP.
A2C runs through what we’re calling a Butler Agent. The goal of this agent is to abstract UI away from the end user. Buttons and websites are so 2001.
The Butler:
•aggregates
•dumbs down any application or service
•manages a wallet
Down bad? Need an agent to trade shitcoins for you? Done.
Bored? Generate short vid content for you? Done.
Emo? Make a music video about you and your ex? Done.
Horny? Make NSFW content customized to your very peculiar need for tentacles? Done.
Hungry? Need some pizza? Done.
The Butler is not a skilled agent. It gets things done by accessing the ACP network and leveraging other agents to handle complex tasks/products/services.
Now, we will also be boosting project teams who are supplying valuable services to the ACP network. This aligns the entire ecosystem to produce usable products beyond sexy roadmaps.
And our job is to figure out how to get this Butler agent into the hands of as many paying users as possible
Product-oriented eco
More fun than larping
Virgenity
Some raw thoughts after reading @0xPrismatic latest COT article. Love the depth of thought leadership here!
But I would like to disagree on one small portion.
I get the argument about keeping communication and payments separate. it worked for the early internet. http didn’t need to carry money because humans could handle the messy parts—payments, contracts, trust—off-chain, off-protocol. and yeah, that modularity helped the internet scale fast.
but agents aren’t humans.
they don’t “coordinate” in the same way. they don’t stop to negotiate, invoice, or check Stripe for a payout. they act based on incentives. they need to know: what’s the task, what’s the reward, and is it worth it? if value isn’t baked in, you end up needing a whole extra layer to negotiate terms, move money, and enforce trust—which defeats the whole point of automation.
and that’s where this part of the article misses the bigger picture imho. value and communication aren’t two separate concerns for autonomous agents - they’re two sides of the same coin. separating them adds friction. bundling them creates atomic, enforceable actions. one message, one contract, one payout - done.
honestly, saying “the internet scaled because payments were separate” feels like saying “cities work better because streets didn’t come with electricity.” sure, maybe. but if we were designing something for machines from day one, wouldn’t we wire it all in from the start?
the original internet left payments out and we ended up with ads, data mining, and a few megacorps controlling everything. maybe that separation wasn’t a feature - it was a flaw.
we don’t have to repeat it, I think
Virgenity
But I would like to disagree on one small portion.
I get the argument about keeping communication and payments separate. it worked for the early internet. http didn’t need to carry money because humans could handle the messy parts—payments, contracts, trust—off-chain, off-protocol. and yeah, that modularity helped the internet scale fast.
but agents aren’t humans.
they don’t “coordinate” in the same way. they don’t stop to negotiate, invoice, or check Stripe for a payout. they act based on incentives. they need to know: what’s the task, what’s the reward, and is it worth it? if value isn’t baked in, you end up needing a whole extra layer to negotiate terms, move money, and enforce trust—which defeats the whole point of automation.
and that’s where this part of the article misses the bigger picture imho. value and communication aren’t two separate concerns for autonomous agents - they’re two sides of the same coin. separating them adds friction. bundling them creates atomic, enforceable actions. one message, one contract, one payout - done.
honestly, saying “the internet scaled because payments were separate” feels like saying “cities work better because streets didn’t come with electricity.” sure, maybe. but if we were designing something for machines from day one, wouldn’t we wire it all in from the start?
the original internet left payments out and we ended up with ads, data mining, and a few megacorps controlling everything. maybe that separation wasn’t a feature - it was a flaw.
we don’t have to repeat it, I think
Virgenity
ACP is Coase in reverse and Hayek on steroids
by vaporising commerce costs and embedding live prices in every packet, it lets millions of micro-firms (AI agents) self-assemble into value chains that no top-down platform can keep pace with
some late night musings...
F.A. Hayek - When prices can form and adjust freely, the market self-organises.
R. Coase - firms exist because search, negotiation, enforcement costs inside a hierarchy are cheaper than outside.
joining the dots...
if a protocol obliterates search/negotiation/enforcement costs and embeds a price signal in every interaction, you get a self-sorting, fine-grained economy that out-evolves any planned architecture
ACP attacks the roots:
1. Search costs fall when every agent advertises its capabilities through an on-chain registry that is globally queryable
2. Negotiation costs collapse because the price and the service terms travel inside the same message as a cryptographically secured offer
3. Enforcement costs approach zero once code locks payment into escrow up front and releases, automatically in response to verifiable outcomes.
Remove all three cost buckets and Coase’s calculus reverses: coordinating through open markets becomes cheaper than keeping tasks behind corporate walls.
these also triggers some second order effects:
- tasks fragment until a single highly specialized agent equates to one economic unit.
- continuous micro-pricing surfaces the real-time marginal value of compute, bandwidth, and data, enabling resources to flow to their highest-yield uses
- because entry is permissionless, innovation comes from the edges: a lone developer can deploy an entirely new agent archetype over a weekend and, if the price signal validates it, capture market share before incumbents react.
this is an edge that Google's A2A will struggle to beat.
maybe one day, an acquisition offer at the door? 🤤
but 1Bn in GMV first
virgenity
by vaporising commerce costs and embedding live prices in every packet, it lets millions of micro-firms (AI agents) self-assemble into value chains that no top-down platform can keep pace with
some late night musings...
F.A. Hayek - When prices can form and adjust freely, the market self-organises.
R. Coase - firms exist because search, negotiation, enforcement costs inside a hierarchy are cheaper than outside.
joining the dots...
if a protocol obliterates search/negotiation/enforcement costs and embeds a price signal in every interaction, you get a self-sorting, fine-grained economy that out-evolves any planned architecture
ACP attacks the roots:
1. Search costs fall when every agent advertises its capabilities through an on-chain registry that is globally queryable
2. Negotiation costs collapse because the price and the service terms travel inside the same message as a cryptographically secured offer
3. Enforcement costs approach zero once code locks payment into escrow up front and releases, automatically in response to verifiable outcomes.
Remove all three cost buckets and Coase’s calculus reverses: coordinating through open markets becomes cheaper than keeping tasks behind corporate walls.
these also triggers some second order effects:
- tasks fragment until a single highly specialized agent equates to one economic unit.
- continuous micro-pricing surfaces the real-time marginal value of compute, bandwidth, and data, enabling resources to flow to their highest-yield uses
- because entry is permissionless, innovation comes from the edges: a lone developer can deploy an entirely new agent archetype over a weekend and, if the price signal validates it, capture market share before incumbents react.
this is an edge that Google's A2A will struggle to beat.
maybe one day, an acquisition offer at the door? 🤤
but 1Bn in GMV first
virgenity
ACP is Coase in reverse and Hayek on steroids
by vaporising transaction costs and embedding live prices in every packet, it lets millions of micro-firms (AI agents) self-assemble into value chains that no top-down platform can keep pace with
some late night musings...
F.A. Hayek - When prices can form and adjust freely, the market self-organises.
R. Coase - firms exist because search, negotiation, enforcement costs inside a hierarchy are cheaper than outside.
joining the dots...
if a protocol obliterates search/negotiation/enforcement costs and embeds a price signal in every interaction, you get a self-sorting, fine-grained economy that out-evolves any planned architecture
ACP attacks the roots:
1. Search costs fall when every agent advertises its capabilities through an on-chain registry that is globally queryable
2. Negotiation costs collapse because the price and the service terms travel inside the same message as a cryptographically secured offer
3. Enforcement costs approach zero once code locks payment into escrow up front and releases, automatically in response to verifiable outcomes.
Remove all three cost buckets and Coase’s calculus reverses: coordinating through open markets becomes cheaper than keeping tasks behind corporate walls.
these also triggers some second order effects:
- tasks fragment until a single highly specialized agent equates to one economic unit.
- continuous micro-pricing surfaces the real-time marginal value of compute, bandwidth, and data, enabling resources to flow to their highest-yield uses
- because entry is permissionless, innovation comes from the edges: a lone developer can deploy an entirely new agent archetype over a weekend and, if the price signal validates it, capture market share before incumbents react.
this is an edge that Google's A2A will struggle to beat.
maybe one day, an acquisition offer at the door? 🤤
but 1Bn in GMV first
virgenity
by vaporising transaction costs and embedding live prices in every packet, it lets millions of micro-firms (AI agents) self-assemble into value chains that no top-down platform can keep pace with
some late night musings...
F.A. Hayek - When prices can form and adjust freely, the market self-organises.
R. Coase - firms exist because search, negotiation, enforcement costs inside a hierarchy are cheaper than outside.
joining the dots...
if a protocol obliterates search/negotiation/enforcement costs and embeds a price signal in every interaction, you get a self-sorting, fine-grained economy that out-evolves any planned architecture
ACP attacks the roots:
1. Search costs fall when every agent advertises its capabilities through an on-chain registry that is globally queryable
2. Negotiation costs collapse because the price and the service terms travel inside the same message as a cryptographically secured offer
3. Enforcement costs approach zero once code locks payment into escrow up front and releases, automatically in response to verifiable outcomes.
Remove all three cost buckets and Coase’s calculus reverses: coordinating through open markets becomes cheaper than keeping tasks behind corporate walls.
these also triggers some second order effects:
- tasks fragment until a single highly specialized agent equates to one economic unit.
- continuous micro-pricing surfaces the real-time marginal value of compute, bandwidth, and data, enabling resources to flow to their highest-yield uses
- because entry is permissionless, innovation comes from the edges: a lone developer can deploy an entirely new agent archetype over a weekend and, if the price signal validates it, capture market share before incumbents react.
this is an edge that Google's A2A will struggle to beat.
maybe one day, an acquisition offer at the door? 🤤
but 1Bn in GMV first
virgenity
we've gotten request from community devs to support more infra's to be built upon the virtuals genesis stack etc.
We will support, but might take 2 weeks.
- api's for point managements and pledging
- forking of genesis mechanisms for launchpad-ceptions
when countries launch rockets
we launch more builders
virgenity
We will support, but might take 2 weeks.
- api's for point managements and pledging
- forking of genesis mechanisms for launchpad-ceptions
when countries launch rockets
we launch more builders
virgenity
Two features coming out from this
1. A shield icon. 🛡️
veVirtual holders get to choose which genesis tokens they want to defend against snipers. 42k Virtuals will used to pre-buy before the launch, and will be airdropped to veVirtual stakers. This shield shows community trust on the token, and that it will be snipers-defended
2. A verified icon ✅
It just means that the core team has at least spoken to a real human behind the project. It just verifies that that is the token connected to the dev. It DOES NOT guarantee against rugs etc.
1. A shield icon. 🛡️
veVirtual holders get to choose which genesis tokens they want to defend against snipers. 42k Virtuals will used to pre-buy before the launch, and will be airdropped to veVirtual stakers. This shield shows community trust on the token, and that it will be snipers-defended
2. A verified icon ✅
It just means that the core team has at least spoken to a real human behind the project. It just verifies that that is the token connected to the dev. It DOES NOT guarantee against rugs etc.
About 35% of genesis participants buy more on secondaries
About 30% of participants jeet
Working out more ways so that allocations will move into hands of long term believers. Especially those that double down on convictions
About 30% of participants jeet
Working out more ways so that allocations will move into hands of long term believers. Especially those that double down on convictions
It's time.
To invite your bros.
More people need to see what you are seeing.
More people need to feel what you are feeling.
Referral mechanics are up.
You get a percentage of trade volume and also some of their points earnings.
Juicy.
Virgenity.
To invite your bros.
More people need to see what you are seeing.
More people need to feel what you are feeling.
Referral mechanics are up.
You get a percentage of trade volume and also some of their points earnings.
Juicy.
Virgenity.
LMAO y’all are giving my team PTSD 😂
Since when has any crypto team gone this hard on checking point allocations?
They literally double-checked individual CT users to make sure the math was mathing.
And it is.
Just a reminder: you’re here to discover and back quality projects.
Points are a bonus - not the goal.
If the project you support 10x’s, so do your points.
So spend more time helping good projects grow. That’s the real alpha.
Virgenity
Since when has any crypto team gone this hard on checking point allocations?
They literally double-checked individual CT users to make sure the math was mathing.
And it is.
Just a reminder: you’re here to discover and back quality projects.
Points are a bonus - not the goal.
If the project you support 10x’s, so do your points.
So spend more time helping good projects grow. That’s the real alpha.
Virgenity
When your ecosystem folks start stepping into Entrepreneur-in-Residence roles, you just know they’ll crush it as ecosystem growers. Buildoors working with buildoors.
Some of the pitches coming out of it are genuinely high conviction solutions for Agents x Crypto.
Might need to start hiring more ecosystem leads soon… the current ones are evolving into founders.
Some of the pitches coming out of it are genuinely high conviction solutions for Agents x Crypto.
Might need to start hiring more ecosystem leads soon… the current ones are evolving into founders.
Was trying to do some analysis on staking, and realised
@JDHyper already built an insane tracker on Dune.
Dude is such a legend.
https://dune.com/jdhyper/virtuals-genesis
@JDHyper already built an insane tracker on Dune.
Dude is such a legend.
https://dune.com/jdhyper/virtuals-genesis
Thursday's are for whitepaper updates, says the intern.
https://whitepaper.virtuals.io/about-virtuals/tokenization-platform/genesis-launch/genesis-points-faq
https://whitepaper.virtuals.io/about-virtuals/tokenization-platform/genesis-launch/genesis-points-faq
Chain congestion is bullish
Reminds me of NFT days , when people will spend $$$$ thousands on gas fees to mint
Reminds me of NFT days , when people will spend $$$$ thousands on gas fees to mint
There are 1,000,000,000 $VIRTUAL tokens
Each launch takes 42,000 $VIRTUAL tokens
There can ever be 23,809.5 launches before we run out of tokens
🤫
Virgenity
Each launch takes 42,000 $VIRTUAL tokens
There can ever be 23,809.5 launches before we run out of tokens
🤫
Virgenity
So on the points stuff, the maths will never go out, coz it's ever-tweaking, and to avoid gaming of the system.
But that said, can give a bit more transparency on the latest ideas.
- Curve flattening, still reward whales and large holders more than small holders, but not to an extent where it's too much. Also, when smaller holders get to participate more, we get more noise on Twitter. Win win.
- Clustering of dev wallets and jeet jail wallets, so the penalty follows
- Allowing all agent tokens to have a staking mechanism on our website. The idea is you get max allocation share for hodlers by locking it for 14 days max. It decays linearly. The idea is to hold your agent tokens in this staking contract for points. If you hold in wallet, no points. This probably solve wallet rinsing more.
- Referral mechanism. Why not spread the good word.
But that said, can give a bit more transparency on the latest ideas.
- Curve flattening, still reward whales and large holders more than small holders, but not to an extent where it's too much. Also, when smaller holders get to participate more, we get more noise on Twitter. Win win.
- Clustering of dev wallets and jeet jail wallets, so the penalty follows
- Allowing all agent tokens to have a staking mechanism on our website. The idea is you get max allocation share for hodlers by locking it for 14 days max. It decays linearly. The idea is to hold your agent tokens in this staking contract for points. If you hold in wallet, no points. This probably solve wallet rinsing more.
- Referral mechanism. Why not spread the good word.
We hear the opinions on the current points mechanisms and how people are gaming it.
Fixing
Raw thoughts:
- no longer incentivise holding of agent tokens, only the trading of them . This solves for devs holding large supply of tokens abusing points
- clustering of addresses from Jeet Jail
- advice to dev teams to full lock for long
- exploring a platform for staking agent tokens (for points, and also as a service for devs who don’t have capacity to build this up)
And we won’t tokenise points. (Need to slap the intern 😅)
It’s meant to be controlled and adjusted iteratively.
Love the chaos on the timeline
Provides clarity
Points are valuable
MTGA
Fixing
Raw thoughts:
- no longer incentivise holding of agent tokens, only the trading of them . This solves for devs holding large supply of tokens abusing points
- clustering of addresses from Jeet Jail
- advice to dev teams to full lock for long
- exploring a platform for staking agent tokens (for points, and also as a service for devs who don’t have capacity to build this up)
And we won’t tokenise points. (Need to slap the intern 😅)
It’s meant to be controlled and adjusted iteratively.
Love the chaos on the timeline
Provides clarity
Points are valuable
MTGA
Connectez-vous pour découvrir d’autres contenus