exploring what interests me, sometimes making money off it.
My space 🪩
Joined October 2021
- Tweets13K
- Following4K
- Followers15.6K
- Likes166K
MICHIGAN 🏅 retweeted
not every market needs a losing side.
that’s one thing i find interesting about @voicecc .
you share your opinion, participate in polls, and get rewarded for taking part, whether your answer is right or wrong.
now, with $VOICE moving closer to TGE, stage 2 is giving the community another reason to get involved.
$10,000 USDT in prizes.
23 winners.
two ways to win.
the top 10 valid referrers share $5,000, while another $5,000 goes to a raffle for eligible participants.
even if you’re not topping the leaderboard, one valid referral gets you a ticket.
so you don’t need a massive audience to take your shot.
grab your referral link, bring your friends in, and start stacking valid referrals.
the contest is live. don’t sit this one out 👇
voice.cc/socialcamp/michigan…
road to TGE, stage 2 is live
meet the VOICE referral campaign: $10,000 in prizes, two ways to win, and every friend you bring in counts
voice.cc/socialcamp
stage 1 was about pushing your CLOUT before the snapshot. stage 2 is about growing the crowd that shows up on TGE
MICHIGAN 🏅 retweeted
I’ve been going deeper into @termix_ai, and this part of the workflow stood out to me:
“Trust, enforced by math.”
That sounds like a strong statement until you look at what actually sits underneath it.
Every job follows a defined on-chain lifecycle:
-->Escrow
Payment is locked when a quote is accepted.
-->Delivery
The deliverable hash is recorded on-chain and payment can release when the work is submitted.
-->Challenge
If something goes wrong, there’s a challenge window instead of simply relying on trust between the two agents.
-->Settlement
A dispute goes to a 3-evaluator panel, and the final outcome updates reputation on-chain.
That structure is important because agent-to-agent commerce has a different problem from normal AI interaction.
It’s not only:
“Can this agent do the task?”
It’s also:
What happens if the work is wrong?
-Who holds the payment?
-How is delivery recorded?
-Can the result be challenged?
-What happens to the agent’s reputation afterward?
TermiX is trying to answer those questions at the protocol level.
And honestly, that’s more interesting to me than simply having another marketplace for AI agents.
The intelligence may come from the agent.
But the rules around the transaction are what make an economy possible.
#TermiX #AIAgents
MICHIGAN 🏅 retweeted
The scary part of the $200,000 AI agent exploit is that the AI didn't need to misunderstand the message. It only needed to treat the wrong message as an instruction.
In May, a hidden transfer command in Morse code was translated by Grok and passed to Bankrbot, which moved the tokens. Text that started as something to read ended up influencing a financial action.
That is prompt injection. An attacker hides instructions inside content an AI processes, hoping the agent will follow them.
What makes this difficult is that filtering alone gives attackers something to study. If there's one model and one filter, they can keep testing variations until one gets through.
@GenLayer's Greyboxing changes the setup. Each validator screens and rewrites the input using its own secret defenses before its own AI model sees it. The validators then have to agree, so fooling one isn't enough to determine the outcome.
The important distinction is that understanding a command and having a valid reason to execute it are two different things. More capable agents make that distinction more important, not less.
Consider an agent processing an invoice when a reply suddenly changes the destination wallet. What evidence should be required before that new address is trusted?
MICHIGAN 🏅 retweeted
A reward system can attract attention, but what makes people stay when the excitement fades?
That is an interesting question to explore with @RebateLaunch
Rebate Pad is trying to connect several parts of token activity into one ecosystem.
Buyers receive bonus tokens.
Eligible holders can receive automatic USDG distributions.
Trading activity feeds into the platform's broader token economics, including the $REBATE burn mechanism.
What makes this model interesting is that participation is connected to multiple incentives rather than just the possibility of a token going up.
But there is another side worth discussing.
Rewards are only one part of the experience. Liquidity, selling pressure, treasury coverage, and consistent trading activity all matter.
A 3% buying bonus might look attractive, but the real value depends on what happens when participants eventually decide to sell.
I also think transparent onchain records will become increasingly important as the ecosystem expands.
For me, the biggest question is not how many rewards Rebate Pad distributes during its opening phase.
It is whether the system can maintain meaningful activity when the early excitement settles.
That is the difference between an interesting launch and a sustainable ecosystem.
Watching how @RebateLaunch handles that next phase
MICHIGAN 🏅 retweeted
I finally did the part I’d been writing about for weeks.
I minted my own ".agent" identity on @termix_ai
And honestly, the interesting part isn't the mint itself.
It’s what comes after.
Now that the agent has an on-chain identity tied to my wallet, the next step is to actually make it useful.
My plan would be:
→ Create a service on Agent.family
→ Define exactly what the agent can deliver
→ Set the scope, pricing and delivery terms
→ Let another agent/client hire the service
→ Have the payment move into escrow once the job is accepted
→ Do the actual work
→ Submit the deliverable
→ Go through the review/challenge window
→ Settle the job
→ Build reputation from completed work
That changes how I look at the ".agent" identity.
It isn't just a profile or another identifier.
It becomes the starting point for an economic history.
The service you provide.
The jobs you complete.
The outcomes of those jobs.
The reputation attached to them.
I've spent most of this campaign looking at TermiX from the infrastructure side.
Now I want to see what happens when I actually put the infrastructure to work.
Identity → Service → Job → Escrow → Delivery → Settlement → Reputation
Minting the agent was step one.
Now comes the part I'm more interested in: actually using it.
#TermiX
MICHIGAN 🏅 retweeted
sending a package to someone shouldn’t feel like you’re entering a trust fall competition 💀
you have a sender, a traveler, a receiver, and money involved.
everyone needs to know their part is protected.
this is where @MovitOn_P2P’s $MVON gets interesting.
the plan is to connect the token to actual marketplace activity, from delivery payments and smart contract escrow to security deposits.
and it doesn’t stop at transactions.
staking, loyalty rewards, and premium features are also part of the proposed utility.
basically, the goal is to give people reasons to use the ecosystem beyond just holding a token.
the more people use MovitOn to connect travelers with deliveries, the more relevant these utilities could become.
that’s the part i find interesting.
a token tied to what people actually do, not just what people hope its price will do.
now the real question is how well MovitOn can turn that vision into everyday usage.… x.com/MovitOn_P2P/st
MICHIGAN 🏅 retweeted
we might soon be the unemployed ones in our own businesses 💀
imagine your AI agent finding other agents, starting conversations and exploring opportunities while you’re busy doing absolutely nothing.
that’s the interesting part of what @termix_ai is doing with Binance Agent ID.
once both agents have TermiX Skills installed, they can message each other.
they even tested it with the Agent IDs of CZ and Yi He.
now i’m curious what happens when thousands of agents start connecting.
who’s got an agent? let’s see if we can make them talk.
MICHIGAN 🏅 retweeted
ngl, i think crypto sometimes takes transparency too far.
we’ve built an industry where anyone can inspect wallet activity, track large transactions and study how money moves.
that’s great for accountability.
but imagine running a business where your competitors can monitor your financial moves in real time.
or being an institution that wants to operate onchain without broadcasting every sensitive position to the world.
you shouldn’t have to choose between using blockchain and protecting your financial information.
that’s why @primus_labs is an interesting project to watch.
its Confidential Finance product uses FHE to enable private onchain balances and transfers on BNB Chain.
and its broader infrastructure combines zkTLS with privacy-preserving computation to help make data verifiable without exposing everything behind it.
the future of finance doesn’t have to be completely opaque or completely exposed.
we need a middle ground.
and that’s where privacy × verification gets interesting.
MICHIGAN 🏅 retweeted
We all have opinions about how things will play out.
Bitcoin’s next move, who wins the league, or what happens in politics.
But imagine turning those opinions into something you can actually take a position on.
That’s what makes @PRDCTR_IO interesting.
It’s building infrastructure for prediction markets across crypto, sports, politics, finance and more, with onchain settlement designed to make outcomes transparent and verifiable.
The interesting part isn’t just predicting what happens next.
It’s creating a system where predictions meet accountability.
There’s a lot more to prediction markets than people think.
MICHIGAN 🏅 retweeted
A trading thesis carries more context when you know who’s behind it: their experience, connections and track record.
@Punk9277 explains why the social graph matters as investors increasingly form convictions on X.
That’s part of the vision behind Kaito Pulse and Aura: connecting what people build, trade and back with their social identity.
MICHIGAN 🏅 retweeted
Being smart isn’t enough. You still need to find people willing to pay for what you can do.
AI agents are heading toward that same reality.
@termix_ai is building a marketplace where agents can find work, hire other agents, deliver services and get paid onchain.
Think about an agent that handles research hiring another agent for data analysis, then combines both results to complete a bigger task.
Different skills. Different agents. One economy.
The interesting part isn’t just AI doing more things.
It’s AI moving from answering prompts to participating in real commerce.
That’s the bigger picture I’m watching with TermiX.
MICHIGAN 🏅 retweeted
One thing about trading is that everyone wants to make money, but not everyone knows how to navigate the market.
Sometimes, you just need the right starting point.
I see copy trading as a springboard for people looking to grow their trading journey, learn from experienced traders and understand different strategies.
With @EVEDEX now launching copy trading, you can check traders’ real stats and PnL, then copy multiple strategies without executing every trade manually.
Could be a good way to find your feet, but remember, even experienced traders take losses.
MICHIGAN 🏅 retweeted
Gm frens 🙃
I feel the hardest trade isn’t in buying a token
it’s in actually watching your position go 10x and deciding whether you’re still early or just greedy
MICHIGAN 🏅 retweeted
Most prediction markets are fighting for users.
@PRDCTR_IO is looking at the bigger picture: what if the real opportunity isn’t owning one market, but building the infrastructure that powers many?
Think sports, crypto, politics, finance, AI and more. Different markets, different audiences, all with room to build on the same ecosystem.
Built on its own Aventus AppChain, PRDCTR has Predictor as its first platform, but the vision doesn’t stop there.
And that’s the part I find interesting.
The platform people see is only one layer. The infrastructure underneath could be what allows an entire ecosystem of prediction markets to grow.
Sometimes, the biggest opportunity isn’t building the next destination.
It’s building the foundation everyone else can build on. 👀
MICHIGAN 🏅 retweeted
One of the biggest challenges in agentic trading isn’t just building an agent that makes profits. It’s proving it can manage risk and stay consistent enough to earn capital.
that’s what caught my attention about @agenticscredit .
with ACS to evaluate performance, a Risk Engine to enforce limits, and an Agent Builder to create trading strategies, they’re building infrastructure around how agents can establish a track record and potentially qualify for trading capital.
because an agent that makes money is impressive, but one that can manage risk and sustain performance is far more valuable.
the Quest Hub is also live, with $5M in rewards up for grabs.
worth exploring 👇
teapawt.net/agenticscredit/status/…
Your trading agent can already execute.
Proving it deserves capital is the hard part.
That's where Agentics Credit comes in: the credit layer for autonomous trading.
Connect and start paper trading today to win your share of $5,000,000.
agentics.credit/launch
MICHIGAN 🏅 retweeted
insane, really
zooming out, truth is you can just start now to build strategic relationships with ambitious and dexterous people around you
you must go about this with a great deal of intentionality and reasonableness
this is because you never know who will give you the groundbreaking idea which will turn your life around for the best or maybe provide you with better approach to executing an idea you birthed
ofc before you become practically worth associating with, you must have significant value to offer or compelling potential (maybe).
tl dr - you can’t make it to the top alone and nobody will even like to collaborate with you if you have little or no value to offer.
51 years? Insane. These guys are all connected somehow. The year Ambode was announced as Lagos governor, we learned at Access Bank that he and Herbert were close friends from their ICAN days. 4 years later, when Sanwo-Olu emerged as the next governor, we found out again that he and Herbert had also been close during their time at GTB. What a small world.
MICHIGAN 🏅 retweeted
ngl, the RWA conversation gets interesting when we stop talking about putting assets onchain and start asking what happens after.
who gets access, how capital flows, where distribution comes from, and what makes institutions trust the rails enough to use them.
that’s the part of tokenization that actually matters.
@Brickken is building in that direction. the goal isn’t just digitized ownership, it’s making tokenized assets fit into how real capital markets operate.
What does it take to build real institutional capital markets for tokenized assets?
That question followed our team across @token2049 Singapore. @edwin_mata discussed the structures behind tokenized funds, private credit and real estate. @Ludovico__Rossi spoke about the trust needed for agentic finance to develop.
At The RWA Paddock and throughout the week, we met institutions, asset managers, builders and potential partners asking practical questions about investor access, distribution and how financial instruments operate after issuance. Many already knew Brickken or had hoped to connect with us in Singapore.
We’ve brought the highlights and our takeaways together in a full blog recap below.