TraiBo19 | Duc 🚢 (@BoTrai9) header
TraiBo19 | Duc 🚢 (@BoTrai9) profile picture

Content Creator | Web3 & Airdrop Sharing opportunities, insights & real-world perspectives Japan 🇯🇵 – Travel | Food | Culture 🇻🇳🇯🇵

Joined July 2021
Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)Photo by TraiBo19 | Duc 🚢 (@BoTrai9)
このアカウントでは日本の暮らし・風景・文化について投稿しています 🇯🇵 (Chia sẻ về cuộc sống & văn hóa Nhật) #暮らしの日本
37
6
1
104
3,761
The “activity” part of Zaps is actually quite interesting. Zaps are tied to trading activity within the ecosystem, rather than simply counting how many transactions a wallet makes. That creates a simple connection: Onchain Activity → Zaps → Leaderboard So the important part isn’t just being active, but how that activity is recognized within the system. I’ll keep testing Zaps and see how this translates into my own ranking.
1
1
13
AUTONOMY NEEDS LIMITS. The way @meetmotes approaches risk management isn't about restricting an Agent from acting. It's about defining clear boundaries for autonomous execution. The control layers include: → Owner-defined limits → Policy enforcement → Transaction simulation → Monitoring → Emergency pause The Agent can still operate autonomously, but its actions must stay within the predefined boundaries. Autonomy is useful when the boundaries remain under control. Is this the right approach for bringing AI Agents into onchain execution? What do you think?
1
1
13
zScore looks at a wallet’s longer onchain history. Zaps focus more on qualifying trading activity happening through the ecosystem. And it’s not simply about making more transactions or pushing more volume. That makes me curious about one thing: What actually makes onchain activity valuable enough to count?
4
1
7
51
FROM IDEA TO ONCHAIN ACTION One thing I’m looking at with @meetmotes is what happens after an Agent decides to act. The flow is more structured than simply signing a transaction: OBSERVE → PROPOSE → SIMULATE → VALIDATE → AUTHORIZE → EXECUTE → SETTLE → RECONCILE The interesting part is the simulation + validation before execution. The Agent can propose an action, but the system checks the expected outcome, policy, exposure and limits before anything goes onchain. For autonomous Agents, that separation could matter just as much as the intelligence itself. Personally, I find this direction interesting. But is Motes building in the right direction for the next wave of AI Agents? What do you guys think? hazels.io/studio/pulse/motes…
12
12
261
TraiBo19 | Duc 🚢 retweeted
🎁 GIVEAWAY TIME! 🎁 🔥20 GTD Whitelist spots for @kindotsui up for grabs 🏆 ⏰ Winners will be randomly picked within 24h. To enter: ✅ Follow @kindotsui & me @jackma1994 ✅ Like & RT this post. ✅ Tag 1 friend and Drop your Sui wallet in the comments. ✅ Join the Kinguard Discord: discord.gg/kinguard Must follow all requirements to be eligible. Good luck everyone! 🔥 #Sui #Giveaway #Whitelist
52
26
1
78
2,059
Tháng này ăn cơm thấy hổng ngon,không khí cũng hổng vui… Thì ra là vậy …😂😂
11
11
145
I’ve been exploring @zerufinance and its Zaps ecosystem. What caught my attention is the idea behind it: Onchain activity can become a measurable signal. ZeruAI uses zScore to read a wallet’s longer-term onchain behavior, while Zaps focus on economic activity happening across the platform. From there, the system connects activity with leaderboards, Clans, campaigns and rewards. So the bigger picture looks like: Behavior → Trust → Activity → Rewards I’m going to spend the next few days testing Zaps and digging into how this actually works. zaps.markets/?invite=Z4j3BRn…
9
4
66
HOW MUCH ACCESS DOES AN AGENT GET? One interesting part of @meetmotes is the use of SESSION KEYS. The idea is simple: An Agent shouldn’t need the same level of authority as the owner. Instead, its session key can be limited by: → Approved contracts & functions → Spending limits → Time duration → Revocation So even if the Agent is acting autonomously, its permissions remain bounded. OWNER = HIGH AUTHORITY AGENT = LIMITED AUTHORITY For me, this is an important distinction when AI starts interacting directly with onchain assets. The question is no longer just “Can the Agent act?” It becomes: “How precisely can we define what the Agent is allowed to do?” hazels.io/studio/pulse/motes…
🤖 Made with AI
4
1
5
51
What if the GPU sitting idle in a data center or at home could become part of a global AI compute network? That’s the basic idea behind Nosana. Here’s how the pieces fit together (1/7)
🤖 Made with AI
7
8
99
And the model works from both sides: 🖥️ GPU providers contribute compute 🤖 Builders consume compute ⚙️ Nosana coordinates the workloads ⛓️ solana:nosXBVoaCTtYdLvKY6Csb4AC8JCdQKKAaWYtx2ZMoo7 is part of the network's economic layer The goal is simple: make GPU capacity more accessible and put unused hardware to work. (6/7)
1
10
What I find interesting about Nosana isn't simply “decentralized GPUs.” It's the attempt to turn distributed GPU capacity into usable infrastructure for real AI workloads. If AI compute keeps growing, the question isn't only how many GPUs we have — but how efficiently we can access and use them. (7/7) #NOSvember #CreatorChallenge
9
500 anh em có vào thử sức không nào? Đùa chứ 2 ngày nữa done rồi, vào ngó chơi thôi mấy ní ..
A new campaign just went live on Rarelist! Join now: rarelist.io/projects/onlystr… @onlystreetsfun is the first launchpad built to turn real-life moments into coins. Duration: Nov 11–17
2
2
126
One thing I’m watching closely is how Rarelist is changing its scoring system. The team has been working on better ways to filter low-quality content and the upcoming campaigns will use upgraded scoring algorithms. That makes the leaderboard more interesting to watch. If quality really matters more than volume, consistent and useful content should have a better chance to stand out. Let’s see how it plays out. 👀 @rarelistio #Rarelist #RARE
1
1
2
38
ONE NFT. ONE AGENT. ONE SMART ACCOUNT. After looking at the Intelligence → Policy → Execution structure, I think the more interesting part of @meetmotes is actually the account architecture behind it. Motes is building around a simple mapping: 1 NFT → 1 AGENT IDENTITY → 1 SMART ACCOUNT The NFT isn't supposed to be the thing executing transactions. Instead, it acts as the identity and ownership reference for the Agent. The Smart Account is where the financial side happens: → Holds the assets → Enforces permissions → Applies spending & risk limits → Validates transactions before execution And the AI layer sits above that account, analysing information and preparing potential actions. That creates an important separation: NFT = IDENTITY AI = INTELLIGENCE SMART ACCOUNT = CONTROL What I find interesting here is that Motes isn't simply trying to make an AI that can use financial applications. It's trying to give each Agent its own programmable financial account, while keeping the owner's rules enforceable at the account level. The architecture makes sense on paper. The bigger question for me is: HOW WELL DOES THIS MODEL HOLD UP IN REAL ONCHAIN EXECUTION? That's what I'll be looking at next. hazels.io/studio/pulse/motes…
1
1
35
TraiBo19 | Duc 🚢 retweeted
A new campaign just went live on Rarelist! Join now: rarelist.io/projects/onlystr… @onlystreetsfun is the first launchpad built to turn real-life moments into coins. Duration: Nov 11–17
2,433
1,148
1,936
3,284
315,978