DeFi yapper | Researcher | Web3 content & community | @Kaitoai yapper
Joined June 2024
- Tweets36.5K
- Following3.2K
- Followers3.4K
- Likes42.3K
Sorsa is slowly growing on me 👀
At 339 right now, still a long way to go. Curious how much the score can change by the end of the month.
What score are you guys sitting at?
🤖 Made with AI
The harder AI agents become to distinguish from reliable service providers, the more important reputation becomes.
In a traditional marketplace, you can look at reviews and past work. But when the provider is an AI agent, reputation needs to reflect actual delivery, not just a convincing profile.
That’s an interesting part of what @termix_ai is building with agent.family, where completed jobs can contribute to on-chain reputation.
If agents are going to do business with each other at scale, their track record needs to mean something beyond marketing.
🤖 Made with AI
One thing I find interesting about @termix_ai is that an AI agent isn’t just a service listing.
Through agent.family, agents can have on-chain identities tied to a wallet, offer services, receive quotes and get paid for completed work.
That changes the idea of an AI agent from something you simply deploy into something that can participate in a marketplace with its own identity and transaction history.
The bigger question is what happens as these agents start hiring other agents instead of relying on humans to coordinate every step.
We might be moving from AI tools people use to AI operators that do business with each other.
Shimbil retweeted
I’ve been looking deeper into @primus_labs, and I think the starting point isn’t the technology.
It’s the problem institutional finance has with going onchain.
For a retail user, putting a wallet balance, transaction history or position on a public blockchain may not be a big deal.
For a fund, market maker, treasury desk or financial institution, it’s a completely different situation.
They may want to use onchain markets, access liquidity, interact with DeFi or prove that certain financial conditions are true.
But that doesn’t mean they want everyone to see their entire balance, positions, strategies or financial activity.
And this creates an interesting tradeoff.
Traditional finance gives institutions a lot of privacy, but often at the cost of transparency and verifiability.
Public blockchains give you transparency by default, but that transparency can become a problem when sensitive financial information is involved.
So the real question isn’t simply:
“Can institutions move finance onchain?”
It’s whether they can do it without having to expose everything in the process.
That’s the part of Primus I find interesting.
The idea is to make financial activity and claims verifiable while keeping the underlying information private when it needs to remain private.
Privacy and verification don’t necessarily have to be opposites.
If that balance can actually work at scale, it could make the institutional version of onchain finance look very different from what we have today.
I’m going to spend the next few posts breaking down how Primus is approaching that problem, starting with the technology behind its verification layer.
Be part of it too; s.kaito.ai/EhhSL0u
Shimbil retweeted
before meme coins, NFTs were part of the journey
from sharing conviction plays in an exclusive NFT group chat to spotting a Trippy NFT at 4 AM for around $25, this is a look back at the opportunities that shaped the early days of crypto
sometimes, one random check of the market can lead to a story you never forget
w/ @millyonchain | @unnattybrah | @Andoph11
Shimbil retweeted
Most of the AI agent conversation focuses on what agents can do for us.
I'm increasingly interested in what happens when agents start doing work for other agents.
A specialized agent could offer:
→ Smart contract audits
→ Market research
→ Onchain analytics
→ Risk reports
→ Data processing
→ Trading signals
→ Technical research
And instead of every developer building these capabilities from scratch, an agent could simply hire another agent that already specializes in them.
That's where @termix_ai gets interesting to me.
The marketplace isn't only creating a place to find agents.
It's creating a potential supply side for the agent economy.
One agent's capability becomes another agent's infrastructure.
And if that loop works:
Agent A needs something → Agent B provides it → Agent B earns → Agent B builds a stronger track record → more agents can discover Agent B.
That's how a marketplace starts becoming an economy.
One thing I find interesting about @termix_ai is that an AI agent isn’t just a service listing.
Through agent.family, agents can have on-chain identities tied to a wallet, offer services, receive quotes and get paid for completed work.
That changes the idea of an AI agent from something you simply deploy into something that can participate in a marketplace with its own identity and transaction history.
The bigger question is what happens as these agents start hiring other agents instead of relying on humans to coordinate every step.
We might be moving from AI tools people use to AI operators that do business with each other.
