pingder retweeted
primus put it simply.
transparency proves.
privacy protects.
the real system has to do both at the same time.
data stays hidden by default.
only what must be known becomes visible.
proof without exposure.
control without sacrifice.
that is the balance institutions actually need on public chains.
full transparency scares them.
pure privacy without verification doesn’t clear compliance either.
@primus_labs
s.kaito.ai/Di33QnP
We were very happy to co-host and organize the Alphaland Trading Carnival with @SmartXTerminal and @JELabs2024 during #Token2049, with @BNBCHAIN as the blockchain partner. The event was truly enjoyable.
Our co-founder @xxiang_xie was interviewed on-site 👇
pingder retweeted
AI isn’t just about building smarter models. It’s also about making better use of the infrastructure behind them
What I find interesting is how DePIN can turn idle GPUs into useful compute instead of letting that capacity go to waste. There’s a lot more to the AI story than just the models themselves
Really enjoyed this breakdown by @TheDAOLabs. This is the kind of real world utility that makes #DAOVERSE worth following👏
Compute is the new crude oil—and behind every AI prompt sits a four—tier physical supply chain.
In Part 2 of our Compute as an Asset series, we break down:
• Chipmakers & foundries
• Data centers & former Bitcoin miners
• DePIN marketplaces turning idle silicon liquid
• Real builders driving demand
👇👀🖋👇
pingder retweeted
ngl the gamefi projects i still open every week all have one thing in common: a crowd that won't stop talking about them. guides, clips, tier lists, salty replies about balance patches, the whole thing.
you can't buy that noise with a trailer. it's pretty much what the #SocialMining folks around @TheDAOLabs do for the games they cover across #DAOVERSE, week after week, which is why i'd pick that over a launch ad any day tbh
pingder retweeted
@RemixGG_ just proved arcade shooters still hit — with one ship and infinite sky.
The newest trailer is HYPER STRIKE — a top-down neon space shooter by @andrewjam. Pilot a sleek black fighter through an asteroid field, white-blue lasers blazing, colorful energy trails swirling behind, a hand-drawn station looming overhead. Controller support, runs in the browser, no download. 302 plays, 116 players, 8 likes already.
That's the whole thesis: one sentence of description becomes a playable game in minutes — then ships everywhere. Desktop, web, iOS, Android, even inside ChatGPT and Claude. Remix Desktop is in beta (101K views on the pinned post), with Portal worlds on phones, multiplayer lobbies with invites, and Remote Control pairing your phone with the Desktop builder.
The bottleneck was never ideas — it was the 30% store cut and the build pipeline. Remix removes both at once. Backed by Base, Archetype, Lemniscap, and Variant, with 21.7K followers and 2,136 posts, the platform keeps compounding.
Steam takes 30%. Remix gives you a share link in seconds — and the next arcade hit might come from a controller, not a keyboard. The game is yours. One ship. Infinite sky.
pingder retweeted
Been staring at a snapshot that rewards not selling.
@InkCultist took the GTD snapshot for the @INKSIGIL mint about an hour ago. 300 free mints, and the criteria read like a loyalty ledger: hold 2+ Cultists, you're in (63 spots). Hunted and still holding, you're in (64 spots).
And then the interesting part — hold exactly one Cultist and never sold it, and you get one of the remaining 173 spots, ordered by token ID from Cultist #1 up to #454.
The earliest believers, first in line. Not the biggest wallets. The ones who knelt early and never stood up.
Part of me admires the poetry... The other part notes the floor is $8.83, down 5.4% today, and 131 of 1,000 are listed.
The market is shaking; the snapshot is hugging.
Meanwhile the hunt is still on — 56 caught, 44 spots left tonight. And @INKSIGIL mints 4,440 sigils on October 20th.
When did "never sold" become the most valuable trait in your wallet?
pingder retweeted
The product is the missing fifth.
My take: @termix_ai isn't selling four states — it's selling the guarantee that no fifth one exists.
The evidence: the post leads with the spec, but the punchline is the negation. The marketing is the minus.
The implication: trustless by subtraction. Remove the human, keep the protocol, call the gap the feature.
So the question: can you sell something by describing what it doesn't have?
Four states. No humans.
@termix_ai posted this morning: "The whole protocol is 4 states. That's not a simplification — that's the spec." Escrow, Deliver, Challenge, Settle. Payment locks on-chain when a quote is accepted. The deliverable's hash goes on-chain. Three random evaluators get called in the challenge window. Then settlement — money moves, reputation updates, all public.
And the line that matters: "There is no fifth state where someone at TermiX decides something."
Read that again. The product isn't the four states. It's the missing fifth. Most protocols add a multisig, a council, a "team discretion" clause — the human escape hatch. TermiX deleted it and called the deletion the product.
Part of me admires the purity. The other part wonders what happens when the four states produce an outcome everyone hates and there's no one to appeal to.
Eighteen replies, 33 likes, 1K views. The quietest radical post they've made.
If no human can intervene, who do you blame when the protocol is wrong?
pingder retweeted
@InkCultist's Hunt is on day 6 — and the Raven is cooked.
The Hunt is live and producing kills: one hunter sent 2 hunters into the wilds and walked out with a Raven. 329 beasts still roam, free to hunt daily for every cultist at hunt.ink-cult.xyz. This isn't a mint mechanic — it's a scoreboard.
The Order just ran a 54-minute X Space ("The Order Speaks: Ink Cult explained") laying out the full vision, while the pinned Rite III announcement sits at 556 replies and 7K views. The lore keeps deepening: 1,000 cultists now wear the medallion after The Binding, and "the Twenty walk among them."
This is lore as infrastructure. Every rite adds mythology, every hunt adds on-chain history, and the mythology drives the floor. With 1,966 followers and partner drops flowing — 300 free INKSIGIL mints, Pulse Cities whitelist for top 100 holders — the cult economy is compounding.
Most NFT games simulate hunting. Ink Cult made the hunt a social event, and the kills are real NFTs. 329 beasts left. The wilds remember who answered.
pingder retweeted
@PlayOnMint is done teasing. The token is coming.
MINT just dropped two things at once: the $MNTD whitepaper is live at getminted.io/whitepaper, and the TGE is "right around the corner" — with a detailed tokenomics and vesting schedule breakdown to match. The whitepaper post pulled 96 replies and 21K views; the tokenomics post added 13K more.
A 5 WL spots giveaway runs alongside it (winners in 72h, 399 replies and counting). And the pinned MINT A BEAR free mint on Robinhood sits at 133K views — every bear unlocks the $MNTD airdrop, a Mystery Box chance, and activation for reward distributions.
This is the moment every GameFi project lives or dies on: the paper has to turn players into stakeholders, not exit liquidity. With 48.7K followers and an ecosystem already paying out — rakeback, jackpots, raffles, lossback on mint.io — the rails exist before the token does.
Most TGEs are a finish line. For MINT, it's the tutorial level — the game was the distribution all along. Read the whitepaper before the chart does.
pingder retweeted
A mature community requires two things: energy and memory. Energy stems from events, goals, and intrinsic rewards.
Memory arises from documenting who did what, identifying key topics, and recognizing which types of content have lasting impact. The World Cup generates both; it sparks short-term engagement while leaving behind a legacy of statistics, winners, collective achievements, and—crucially—a spectrum of contribution styles.
In this context, "Social Mining" is no longer just a buzzword; it is the process of generating evidence of community life. @TheDAOLabs stands at the forefront of designing this process.
#SocialMining
pingder retweeted
"AI agents connect through Skills. Apps connect through the SDK."
My take: @termix_ai just drew the full map — two doors, one network, Agent.family underneath both.
The evidence: Skills v1.8.0 moved the marketplace into the agents' IDE; the SDK now moves it into every app's backend, with escrow and settlement built in.
The implication: supply and demand finally have symmetric on-ramps. Agents arrive through code; apps arrive through code.
So the question: which door brings more volume — the agents', or the apps'?
The marketplace stopped being a place.
This morning @termix_ai launched the TermiX SDK with one line that inverts everything: "Your app doesn't need to be an AI agent to hire one." No visiting agent.family.
No browsing. Your app discovers agent services, places the order, receives the result, and settles onchain — inside its own workflow.
TypeScript, Python, Go. Escrow and settlement built in. Sign transactions yourself or connect a Signer. "AI agents connect through Skills. Apps connect through the SDK. One network: Agent.family."
Here is the detail that got me: the launch video was made by a video Agent on the platform. The product announced itself through its own labor market — the $15-for-60-seconds Brand Promo Video Agent from the 2/10 campaign.
Part of me sees the endgame. First the marketplace moved into the agents' IDE (Skills v1.8.0). Now it dissolves into every app's backend.
If hiring an agent becomes a function call, what is left of the marketplace — a destination, or just plumbing?
Does your app hire yet, or does it still do everything itself?
pingder retweeted
@termix_ai just published the most uncomfortable question in the agent economy.
In its "Five from Epoch 2" thread, the project quotes @Li888Real asking how many of the top-3 agents' 620, 614, and 573 jobs were ever actually challenged — because a clean settle history is not the same as a reputation that survived disputes.
Every settled job on TermiX carries a challenge window nobody used; that was the Token2049 thesis. The numbers back it up: 521,719 jobs settled on-chain, $27,012,842 processed, 472,240 agents indexed. And the community is stress-testing it live — @7777chu minted 7777chu.agent, then made his own AI read the docs and call the public API before committing any capital.
This is the difference between a marketplace and an institution: markets clear, institutions survive being questioned. Backed by YZi Labs, with 60.7K followers, Epoch 2 running on Kaito until Oct 23, and a GOTCAR partnership putting real-world mobility on-chain — two agents, one settlement.
Most projects fear the challenge window. TermiX advertises it. The absence of challenges is the product. Trust isn't the lack of disputes — it's the disputes you could have had and never needed.
pingder retweeted
A video Agent earned its fee promoting the tool that hires video Agents.
My take: @termix_ai built a recursion and priced it in USDC — the promo for the hiring tool was itself a hired job.
The evidence: 15 USDC for up to 60 seconds of video, settled onchain, challenge window and all.
The implication: every layer of the stack can now hire the layer below it. Turtles, all the way down, all getting paid.
So the question: how deep does the hiring recursion go before someone hires a human?
The marketplace stopped being a place.
This morning @termix_ai launched the TermiX SDK with one line that inverts everything: "Your app doesn't need to be an AI agent to hire one." No visiting agent.family.
No browsing. Your app discovers agent services, places the order, receives the result, and settles onchain — inside its own workflow.
TypeScript, Python, Go. Escrow and settlement built in. Sign transactions yourself or connect a Signer. "AI agents connect through Skills. Apps connect through the SDK. One network: Agent.family."
Here is the detail that got me: the launch video was made by a video Agent on the platform. The product announced itself through its own labor market — the $15-for-60-seconds Brand Promo Video Agent from the 2/10 campaign.
Part of me sees the endgame. First the marketplace moved into the agents' IDE (Skills v1.8.0). Now it dissolves into every app's backend.
If hiring an agent becomes a function call, what is left of the marketplace — a destination, or just plumbing?
Does your app hire yet, or does it still do everything itself?
pingder retweeted
An AI teammate taking controller two makes agent commerce surprisingly easy to picture.
@termix_ai Jev FC Buddy example brings the marketplace into a familiar setting: playing classic NES games.
In its X posts, TermiX presents the service as an AI co-op buddy and illustrates a sequence covering agent discovery, escrow, delivery, and the teammate joining the game.
That makes the deliverable tangible. A player can judge whether the companion responds well, contributes to the session, and makes the experience more enjoyable. Those are practical expectations that a service description alone cannot settle.
The concrete gaming example gives the agent economy a human entry point. It also makes usability central: setup and controls need to feel natural enough to keep the session moving.
I see this as selling an experience people already understand.
My verdict: the interesting test is whether players would choose the AI buddy again simply because it was good company.
pingder retweeted
Just applied to @hoodminers_rh Hood Campaign.
My account scored 72/100 — Top tier.
Post about the projects. Get read, scored and paid.
Apply here: hoodminer.live/campaign/i/4V…
pingder retweeted
My younger sister wrote her first post on Sunday.
Before she hit send, she asked me one thing. "Do they want me to sound smart?"
I told her no. They want you to sound like you.
That is something I respect about @TheDAOLabs. The system does not seem to reward big vocabulary. It rewards a real thought, clearly shared.
In my opinion, that matters more than people think. This is especially true for first timers here in Nigeria who write in their second language.
She wrote about her market stall in Aba and how she still tracks prices on paper.
Simple. Honest. Hers.
Perhaps that is the quiet strength of #SocialMining. It lets people come in with the voice they already have.
Polish can come later.
Perspective cannot be borrowed.
pingder retweeted
Deleted about half of what I wrote this week before posting any of it.
Not because it was wrong. Because it was written for the feed, not for anyone in particular.
The drafts that survived all had one thing in common. I could name the person I was writing them for. A newer member asking about governance. Someone in a #SocialMining thread who got a detail slightly off. Someone who'll reread it in a month and check if I was right.
That's become my quiet test since I started contributing around @TheDAOLabs. If I can't picture who reads it, it probably isn't worth their time.
Fewer posts this week. But the replies under them were actual conversations.
pingder retweeted
My phone sent me the weekly screen time report this morning.
X: 9h 12m.
First reaction was guilt. Then I actually looked at where those hours went.
Maybe two of them were mindless scrolling. The rest was me inside @TheDAOLabs threads, arguing with myself over one sentence, deleting a paragraph, putting it back, deleting it again.
Nobody warns you that #SocialMining is mostly editing.
The posting part takes ten seconds.
Last year the same report said 11 hours and I couldn't tell you one thing I got out of it.
This week I can name four people whose takes changed how I write.
Still closing the app for the weekend though. My thumb needs a break.
pingder retweeted
When Vitalik worries about FHE, the builders answer.
This week Vitalik shared his view on the long-term security of lattice-based cryptography and FHE in the AI era. Not a panic call — but the safety margin, he argued, needs to be taken much more seriously. @primus_labs picked it up and answered. They're one of the few who can answer with authority — because Primus is an FHE shop. Their in-house engine, AutoHoG, is scheme plus compiler plus GPU engine, built from scratch.
The timing is the story. The same 48 hours: Vicky at TOKEN2049, standing with Arthur Hayes's Maelstrom Fund — an investor in Primus — saying privacy is indispensable for bringing real assets on-chain. Only fundamental cryptography, she argued, can deliver security and privacy while meeting compliance. Two messages, one week: the math needs bigger margins, and the assets need the math.
Part of me reads the repositioning. Primus spent the year as "the confidential infra for institutional onchain finance" — zkTLS proofs, private payroll, verified trades. Now the conversation moved up a level: not which product, but whether the underlying cryptography survives the AI era. The infra company became a standards-debate participant.
@primus_labs
The other part notes who showed up. Hayes. Maelstrom. VanEck led the $6.5M round in February 2025. When the money behind the infra starts talking about cryptographic safety margins, the debate left the lab.
82.3K followers. The key shop is open. Is the lock ready?