Radu | Linx

Teams that innovate, deliver, push boundaries -- in the long run, they are the ones worth paying attention to.

Some of 2026's nastiest exploits had nothing to do with smart contracts. The contracts were absolutely fine. Several attacks were traced back to the off-chain infrastructure that nobody thinks to audit, like: ▶️ RPC nodes ▶️ Relayers ▶️ Indexers These are all in scope for an attacker, but none of them are in scope for your audit firm. That's a problem. Fortunately, @RaduC22 built @linx_labs on @alephium specifically to reduce this surface. Now, he shares exactly why it matters. 🔗 alephium.org/news/post/the-…

If you're interested in deeper technical discussions on this subject or want to provide feedback, my DMs are open. Also for next time I'll try to provide a drawing of the rough bridge architecture so it's easier to reason about it. The first milestone is getting closer which is an alpha testnet version, likely for either BTC or DOGE. Back to work now.

Sadly waiting that long is far from ideal UX if you want to use this bridge, so we had to scrap that approach. Instead we designed a new derivation solution which relies solely on the Alephium address. It involves computing a custom derivation path from 6 SHA256 chunks of the Alephium address itself together with some other fixed components. This approach is reorg safe and the deposit address is available in a few seconds as opposed to waiting for 30 minutes. /5

An update from that first post, it's a bit technical, but will try to keep it straight-forward. When you bridge funds from say dogecoin:native to Alephium, you need to first deposit your DOGE tokens to an address where they will be locked. /1

@nael_idrx Congrats on the new office!

16 months ago it was just screens, coffee, and belief. now it’s a real office filled with people building the future of Rupiah stablecoin

There are still many questions to be answered down the road, but I am excited for what this could bring to @linx_labs and @alephium Also let me know in comments what do you think about this initiative, what potential problems do you see (and what are the solutions 😀). And if you'd like us to build this "in public", this would be more granular updates, some very technical, as we make progress. And thanks for reading until here!🫡 /End

The short answer is we need to bridge those tokens over to Alephium. And that's quite challenging since those are all chains without smart contracts. But definitely not impossible. This week made the first steps: we have a repo, figured out deposit address derivation, connecting to a chain, streaming blocks and detecting deposits. /4

Wanted to share some insights into what I meant by "new collateral types". Will be a thread. Most used collaterals in DeFi lending are ETH, BTC and recently more and more RWAs in loops. On Linx we have the L1 token, $ALPH and bridged $WBTC from Ethereum as the 2 collaterals that can be used to borrow. /1

stfu and ship

@CL207 Similar vibes. Sell in May and go away still a thing?

stock indices literally up only, starting to look not real, almost uneasy feeling, usually i start to feel this much sooner than an actual blow off top tho

@iWalkingCat @_iamsraju Thank you!

@raduC22 @_iamsraju Nice job! 🔥

91% utilization means borrowing demand is outpacing supply. Growing from here means either more ALPH holders depositing, or new collateral types that bring in new borrowers. We'll be working on the second one.

@gametheorizing @naval Yup, it’s easily a 10x if not more, at least for engineering throughput. I feel like we’re extrapolating all the roles/skills that AI can cover though. Like someone said, you can outsource thinking but not understanding.

@raduC22 @naval Yea but imagine the delta of what 10 people can do with the new tools if in a strong loop together, compared to just the obvious delta of the 1 person

The era of vibecoding has gotten a lot of people tapping into some primal 'builder' mentality, they are finding a life force hidden in their evolutionary genetics. Listen to the latest @naval podcast for example on vibecoding. I even speak to some almost-retiree age coders, and they are fired up wanting to ship some new Fomrula1car quality stuff they dreamt of their whole careers and think they can now code up in 6 months. Their epigenome is waking up, their mitochondrial factories going into overdrive from a hibernating life force-- like a chinese factory boss yelling to its minions that NOW is the time to deliver full force and meet that deadline. Seeing how much of their productivity is being unlocked, they compare what is newly possible to the current state of the world - and they can clearly see titillating commercial opportunities. But as we learn from trading markets, you should always think where the puck is going. All the immediate things that are unlocked for you are also now being unlocked for everyone else. If its suddenly easy to make "x application" then by the time you build it, the commercial value will be short lived as the world will have made "x" a commodity already. Any lone wolf out there will be able to compete on the things that are becoming possible for a single person to produce. And there are a lot of smart wolves out there. The real alpha I believe will instead be in lean organizations with leaders able to take groups of ~5-10 wolves and create an aligned incentive structure for them where they prefer to collaborate together in a small pack than be alone. 1. People innately have special talents in different areas, so being able to combine into a team lets specialization occur which unlocks higher quality output. 2. Multithreading people allows for quicker iterating, and less chance of blind spots getting through. 3. Simply, compare how many driven/aligned team of 5-10 superstars there are and how much harder it is to get to that organizational level, than how many solo talented people are out there. Many dont have even the social skills to work with others, its a common side effect of their other superpowers. Bottom line- while the AI tools are unlocking a lot more possibilities for what a single person can produce, the real alpha is in creating a type of organization that can align several of them together. Another barbell we see- hard skills supercharged, and soft skills like understanding psychology and aligning incentives, are what will win. Booksmarts are dead.