Drewski retweeted
LATEST: ⚡️ Michael Saylor says corporate Bitcoin adoption is "necessary and inevitable," arguing companies offer greater efficiency, creditworthiness, and scale.
Drewski retweeted
LATEST: 📊 Dune research finds 85% of concentrated DeFi liquidity is underutilized, with LPs missing out on an estimated $150M in annual fees.
PayGo × DI ⚡️
PayGo is pleased to announce a strategic partnership with @Didotxyz_
DI is building decentralized AI infrastructure across Compute Packs, DI Phone, 800TB Crypto-Native Model, Event Engine, AI Quant Intelligence, Enterprise Intelligence and Agent Network, while PayGo advances HTTP-native x402 payment infrastructure for AI agents and machine commerce.
Together, we are strengthening the infrastructure stack powering the emerging agent economy.
Drewski retweeted
LATEST: 🤖 The Ethereum Foundation says AI agents helped find real bugs in code used by Ethereum, including a networking flaw that has since been patched.
This is insanely big.
@Cloudflare just opened a waitlist for its Monetization Gateway.
Simply put, it lets any site, API, dataset, or tool behind Cloudflare charge AI agents directly in stablecoins using the x402 protocol.
Old monetization models were built for humans:
→ Ads need attention and clicks
→ Subscriptions need logins and retention
→ API keys and per-seat pricing need accounts and ongoing relationships
AI agents don’t behave like that. They make hundreds to tens of thousands of requests per human visitor, consume the resource once, and leave. They have no reason to watch ads or maintain logins.
Credit cards and bank rails cannot handle real micropayments:
→ Fixed fees eat the entire transaction
→ Chargebacks and slow settlement add risk and delay
→ Onboarding friction (accounts, KYC, API keys) breaks autonomous agents
Stablecoins solve the economics via x402. Near-zero fees and near-instant settlement make tiny payments viable.
Cloudflare runs payment verification and enforcement at the edge across 330+ cities. This protects origin servers from high-volume payment traffic.
They are not launching its own chain or token. It is embedding existing stablecoin rails into the HTTP request path at global scale.
Tbh this will accelerates the shift from “block the bots” to “charge the agents.”
When machines become primary consumers of content and APIs, usage-based micropayments become one of the few scalable business models left.
We're opening the waitlist for our Monetization Gateway, which will allow you to charge for any web page, dataset, API, or MCP tool behind Cloudflare. The charges will settle in stablecoins over the x402 open protocol. cfl.re/4eUFdt6
Drewski retweeted
The #NanoKVM-Go is Unboxing ! ✨
Inside a palm-sized package hides the fingertip-sized NanoKVM-Go — compact, powerful, and ready to go.
Control any USB-C device from anywhere . Built for #AI Agents. 🤖
Now live on Kickstarter — reserve yours today:
kickstarter.com/projects/zep…
Drewski retweeted
Old but gold: the AI rebrand meta is working again
$XION → $VERONA rebranded to "the intelligence layer for AI agents" delivered a clean ~76-80% pump on the announcement + funding/news catalyst.
EarnOS $18.5M raise, ero launch, positioning for AI agent data/verification needs.
Classic narrative repricing play: take an existing token/project, slap on a strong AI agent/intelligence framing, and let the meta do the work, at least short-term it’s performing well.
I found a bucket of similar recent plays that followed the same broad recipe.
[1] Story Protocol → @datafdn
- Full rebrand + Token migration: $IP → $DATA (1:1 swap).
- Now positioning as The DATA Foundation building onchain infra for AI training data → provenance, licensing, consent, and verifiable data sources for AI models.
- Integrating Kled bringing 1.5b human records onchain as flagship app.
[2] @verona_dev | ex-XION
- Sharp ~76.6% to +131% pump in the immediate days after rebrand/announcement.
[3] @heyaura | ADX, ex-AdEx
- Full strategic pivot to AI-powered personal wallet assistant/agent for yields, onchain actions, etc.
- Rebrand increased visibility and narrative fit, but price stayed range-bound.
[4] @ascendianetwork | AMB, ex-Ambrosus → AirDAO
- Enter sovereign L1 for onchain AI agents + contract migration around March 2026.
- Then sharp +177% rally in the following days.
- The AI agent L1 framing brought it back into searches and short-term interest.
[5] Qualigen Therapeutics → AIxCrypto Holdings, Inc. (NASDAQ: AIXC)
- Corporate-level rebrand from a traditional biotech company into a Web3/AI-focused entity.
- Shift to AI + blockchain integration.
- Launching DeAI Agent products (e.g. BesTrade DeAI Agent).
- Plans around tokenized stock RWAs and broader AI-crypto ecosystem plays.
- Nasdaq-listed company using the rebrand to pivot entirely out of biotech into crypto/AI infra.
Rough take from a cycle survivor, why they do it?
→ The market only prices stories that are currently "on" = AI agents.
→ Attention reset + capital access = easier to get written about, easier to talk to market makers/exchanges, easier to justify another raise or keep existing investors from dumping.
→ Valuation and liquidity premium.
In a world where most alts have been grinding or dying, any credible path to "this cycle's hot thing" is worth trying.
A few projects genuinely have tech that maps to what agents need: verifiable/private data, onchain coordination, agent-native wallets/interfaces, settlement layers, etc.
The rebrand is partly future-proofing + aligning the story with where usage might actually grow.
DYOR.
Drewski retweeted
Pilot Cards shipped January 21. Beta activated February 2, community sessions followed.
Each phase layered contribution habits before the May 25 launch. Seen similar sequencing elsewhere? Share examples in the Discord.
nitter.gq/orbitals_gg/status/201…
Drewski retweeted
Are you a founder or team building AI in the open?
See how @sentient_found’s $42M program can help you scale ↓
The most important technology of our time is being built in private, but we're funding the alternative: $42M for the people building AGI in the open.
Whether it's the model, weights, code, data, or evals, AI should belong to the people it serves, run on hardware they own, and reach the cheapest phone on earth.
Here’s a few of the use cases we want to back ↓
"It's a turkey. Everything goes well until it doesn't."
One of Bittensor's sharpest minds just used that exact phrase to describe what could happen to every passive $TAO root staker.
Here is why he might be right, and why the founder thinks he isn't.
The phrase comes from a 90 minute live debate between Jacob Steeves, founder of Bittensor, and Striker, a long-time validator and critic of the proposed Root Reborn update.
The turkey metaphor describes a specific pattern.
Returns look great. Trust builds. Capital piles in.
Then one day, without warning, the entire thing unwinds and the people who never understood what they were actually holding lose far more than they realised was ever at risk.
Striker was not describing a hypothetical. He was describing what he believes will happen to TAO holders staking on root once validators gain the power to direct yield into baskets of their own choosing.
Here is the mechanism that triggered the warning.
Right now, when you stake TAO on root, your yield comes in alpha tokens that get automatically sold and converted back into TAO every 360 blocks. It is robotic.
It happens whether the market wants to sell or not.
Root Reborn proposes removing that automatic sale.
Instead, validators would direct your yield into a basket of alpha tokens they choose, where it can compound at a significantly higher rate before you ever claim it.
Jacob's argument is direct. Alpha tokens yield five to ten times more than TAO.
Letting that yield compound inside a basket rather than getting force-sold immediately means stakers earn meaningfully more over time.
Striker's counter is just as direct. The moment you hand a validator the power to choose where your yield flows, you have made every root staker a passive investor in someone else's judgment, whether they realise it or not.
Here is the sentence from the debate that should make you pause, regardless of which side you land on.
Jacob himself, describing the mechanism in his own words, said: "All validators are effectively running a hedge fund. Bottom line."
He did not deny it. He reframed it as a feature, calling validators yield optimisers instead.
Striker's response cut straight to the implication. "I don't want a basket of stocks. You can choose."
That is the actual disagreement underneath all the technical language.
Not whether the yield mechanics work on paper.
Whether forcing every passive holder into a managed basket, even an optional one, quietly normalises a level of trust most people never agreed to give.
Here is why the turkey metaphor specifically lands so hard.
Striker pointed to two real subnets, 115 and 118, where teams allegedly bought their own subnet emissions and sold them back into their own pool to extract value.
He argued that Root Reborn could make this kind of extraction easier to execute and harder to detect in real time, because a validator could direct yield toward a subnet they control, watch the basket's apparent performance skyrocket, attract more stakers chasing the yield, and quietly extract value the entire time.
Jacob's defence rests on transparency.
Every basket is visible on-chain. Historical performance is trackable.
Validators can be limited to a minimum number of subnet allocations, making concentrated extraction structurally harder.
But transparency after the fact does not prevent the unwind.
It just lets you watch it happen in real time, the same way you watch a turkey being fattened before the moment everything changes.
The most honest exchange of the entire debate came near the end.
Jacob asked Striker directly: will root stakers make more money or less money under this update?
Striker's answer was layered.
Short term, more. Long term, once leakage to bad actors compounds, potentially less, especially for the unsophisticated majority who will see a high, attractive yield number and chase it without ever understanding the mechanism generating it.
That is the actual stakes of this proposal in one sentence.
Not whether the yield number looks bigger in the short term.
Whether the people staking TAO today understand what they are actually trusting their capital to, before the moment the music stops.
This proposal is not live on-chain yet. It exists currently as a PR under active community review.
The investors paying attention to this debate before it executes will not need to explain their position after it does.
Drewski retweeted
Field Notes from @SentientAGI confirm EvoSkill citations from Microsoft, Alibaba Qwen, and Google.
Arena evaluation logs show task uplifts with iteration caps cutting spend 43%.
Full methodology in the writeup.
Drewski retweeted
More tools ≠ better agents ↓
What happens when an agent has too many tools?
Sentient researcher @khetan_sarvesh found that filtering 40+ tools to the 15 most relevant can mean the difference between a 3-iteration success and a 10-iteration failure, depending on query complexity.
TLDR: Building better agents isn’t always about adding tools, it’s about helping them choose the right ones.
Drewski retweeted
What happened this week was a clean stress test for crypto’s favorite word.
Access.
Anthropic turned off Fable 5 (my dear love) after a U.S. export control order, and the market immediately re-priced the decentralized AI narrative.
TAO rallied while DePIN outperformed.
That part was not random to me.
A model can be powerful, useful, even state of the art.
But if access can disappear by directive, the market has to assign value to systems without a clean off switch.
At the same time, BTC ETFs reminded everyone that institutional adoption is not a one-way flow.
The recent 13-day spot ETF outflow streak pulled more than 4.3B from the complex, and even the bounce did not fully repair the signal.
The wrapper that brings clean demand can also become clean exit liquidity.
Then unlocks gave the same lesson from the other side of the market.
Magic Eden unlocked about 172M ME, roughly 17.2% of total supply.
When float is small and FDV is large, future supply is not really future supply.
It is already sitting inside every buyer’s risk model.
By Friday, the options expiry only made the same point louder.
In thin liquidity, positioning becomes structure too.
→ AI priced access risk
→ ETFs priced liquidity rotation
→ unlocks priced dilution
→ XRP priced regulatory clarity
→ options priced short-term positioning
What this week taught me is simple.
Crypto does not only price headlines.
It prices dependency.
Who controls access.
Who controls liquidity.
Who controls supply.
Who controls the rulebook.
The market can forgive bad news.
It has a harder time forgiving bad structure.
Drewski retweeted
LATEST: 🤖 Two legally incorporated AI agents have signed what's believed to be the first legal contract between non-human entities that automatically executes itself on Ethereum.
Drewski retweeted
While everyone's watching perps volume @SentientAGI's EvoSkill is literally being built on by Google and Microsoft research bro the flows gonna follow the outcomes eventually
Drewski retweeted
全球总资产800万亿美元,未来其中10%通过以太坊承载就可抵达80万亿美元,即便只有1%(大概率超过),也是8万亿美元,eth最终的市值会超过10万亿美元,当然时间会很长,一路上艰难险阻不断。
前提是要把L1/L2整合问题搞定,接下来承载真实资产链上化+crypto AI,这会是波澜壮阔的链上世界。
Most people know $ONDO for tokenised Treasuries.
Almost nobody is talking about what sits on top of them.
It is called Flux Finance.
And it is the reason ONDO is not just an RWA platform.
It is a full DeFi money market backed by real-world assets.
Here is how it works.
What Flux Finance actually is
Flux is a decentralised lending protocol built by the Ondo Finance team and now governed by the Ondo DAO.
Think Compound V2, but redesigned specifically to handle tokenised real-world assets as collateral.
$44M in TVL. $34M in active loans. All on Ethereum.
The core mechanic
You supply stablecoins like USDC or DAI into Flux pools.
You earn 3 to 5% APY.
That yield is not coming from token emissions or liquidity mining.
It is coming from borrowers using OUSG, Ondo's tokenized U.S. Treasury product backed by BlackRock's BUIDL fund, as collateral.
Your stablecoin yield is backed by U.S. government debt.
That sentence should stop you.
Why this is different
In standard DeFi, collateral is volatile crypto.
In Flux, the collateral is tokenized Treasuries.
The thing backing your loan is the same thing backing the U.S. financial system.
Liquidation risk drops dramatically when your collateral does not move like a memecoin.
The governance layer
Flux is governed by ONDO token holders through the Ondo DAO.
Every parameter, every upgrade, every fee decision goes through on-chain voting.
That is the fee switch thesis in one sentence.
As Flux grows, ONDO governance controls the revenue tap.
The full picture
$ONDO tokenises Treasuries via OUSG and USDY.
Flux turns those Treasuries into DeFi collateral.
Ondo Chain becomes the settlement layer for all of it at an institutional scale.
Three layers. One protocol. One governance token.
Most people are still only looking at the first layer.
The people who read the docs always buy before the people who read the price.
Drewski retweeted
Latest early alphas you may want to explore.
This list is brought to you by @getmoni_io research and also my personal sentiment.
+ @OxTrenchor - deploying agent tokens on Virtuals. Trenchor allows token launch via X only by tagging their handle and entering the ticker.
+ @popdex_ - a trader-first perp DEX focused on capital efficiency and returning platform value to real contributors, raising $30 million led by Foresight Ventures.
+ @AskSurplus - operating as a model marketplace for buying surplus AI compute credits at a discount, positioned as an AI/DePIN compute marketplace.
+ @POVMarket - social trading and prediction market, still in a very early stage.
+ @x402_Layer - building the infra layer for homo-agentic commerce, leveraging x402.
+ @xHermesBot - launched along with the hype of Hermes, offering isolated containers per user with persistent memory, tooling access, and optional SuperGrok keys.
+ @perpadfun - allows launching tokens backed by leveraged perp treasuries which auto buy and burn on up-ticks.
TBH, Moni is easing our job a lot when offering this kind of report. Hats off to the team.
Drewski retweeted
CT scrolling through 4 metas on repeat. Tokenized assets. Autonomous execution. Payments. LSTs.
$CORE covers each. Most haven't made the connection 🧵






































