Ok I just read through vitalik’s paper and if the vision he outlined is executed the implications are absolutely insane especially for the app layer. The paper is basically Eth’s plan on how they will enable computation off chain with Eth verifying the result (ensuring computation occurred and what it did etc etc). Now what this means is the definition of “onchain” itself changes, because instead of every part of an app having to execute inside a smart contract you can push a huge amount of complexity elsewhere and still inherit ethereum’s guarantees, this really does open up a huge amount of design space especially for DeFi. For example with lending you could have really sophisticated systems analysing collateral, liquidity, borrower behaviour and market conditions offchain, with multiple solvers or execution systems competing to produce the best outcome, like the cheapest liquidation route, best refinancing rate, best collateral swap or best way to match borrowers and lenders all while the protocol only accepts outcomes that satisfy its rules. So basically a lot more of the smart stuff behind a financial product can happen outside the smart contract, while Eth still acts as the final layer that checks the rules were followed and makes sure the money moves exactly as it should. The MEV implications here are quite interesting too because more economic activity happening before settlement means more value moves upstream into who gets the information first, who computes the best outcome and what gets presented to Eth. But the thing I’m most bullish on is what this does to the TAM of Eth just because this vision means ethereum doesn’t have to be fast enough to do everything itself it just has to be able to verify everything that matters. And that’s an enormous difference. It’s literally a world computer.

Man I wrote something a couple of days ago about how I think the next financial war is going to be fought over the capital routing layer and this is another pretty big signal of where things are heading. I think we’re going to see the frontier AI labs get a hell of a lot more involved in crypto and DeFi than people expect, I mean @sama positioning here is super interesting. You’ve got @OpenAI building the agents while @worldnetwork , which he cofounded, building the identity infrastructure those agents could use to interact with financial institutions, world already integrated @coinbase x402 and is moving into stablecoin payments, so there’s clearly a lot more going on here than just protecting banks from deepfakes. I expect to see way more involvement from AI labs in DeFi and finance as a whole just because the reality is that whoever controls the identity and trust standards financial institutions accept will end up having enormous influence over which agents get access to capital in the first place. So yeah tldr is that we’ve been massively underestimating just how intertwined AI/DeFi and tradfi is going to become, I don’t see a world where these guys are going to be content just providing the intelligence while everyone else controls where the money goes. Which means our job is to make sure that as this war for control of the capital routing layer heats up, DeFi becomes the financial infrastructure everyone is competing to route capital into. I actually think the frontier AI labs could become some of the biggest drivers of DeFi adoption we’ve ever seen and we’ve got a pretty insane opportunity here if we play our cards right. Let them fight with tradfi over who controls the flow of capital and no matter who wins we should be making sure it all flows onchain lol.

Man I really think the next financial war is going to be fought over the capital routing layer and if you look at what some of the big boys in finance and tech have been doing recently it’s obvious we’re already watching the early stages of it play out. S&P have been making some super smart moves into DeFi, positioning themselves to become one of the firms institutions and AI agents rely on when deciding where to put serious money onchain, they’ve also made credit ratings accessible to agents via MCP, meanwhile you’ve got JPmorgan restricting agents like muse from accessing customer accounts and openAI getting into agent payments and looking into acquiring Trail of Bits. So today your bank holds your money, you decide where to put it and financial products compete to attract you, but when agents start making those decisions the company holding your money, the company operating your agent and the firm deciding which risks are acceptable can all be completely different businesses, all fighting over the insane amount of money to be made controlling the flow between them. The good news is this is insanely bullish for DeFi because we’ve already built financial markets that machines can interact with directly without needing to negotiate access with every protocol, but it also means the customer protocols are competing for is going to change from humans to machines that can compare thousands of lending markets and move money around in seconds, which is a whole different ballgame. I mean think about what that does to the economics of distribution, if you’re operating agents that direct billions in capital you’re in a really powerful position. You’ve got lending protocols competing for access to your flow, you can negotiate better terms, take a cut of the capital being routed and build an absolutely enormous business without ever having to provide liquidity or take on any credit risk yourself. But that’s mainly looking at things from the retail side, institutional capital is a different beast because even if an agent finds the best lending opportunity in DeFi, it still has to adhere to the risk mandates of whoever owns that money. And that’s why I think S&P’s moves are so smart because if their DeFi assessments become part of the criteria institutions use to decide where their agents can allocate, they’re basically positioning themselves to influence which onchain businesses get access to some of the biggest pools of capital in the world. That also creates some pretty crazy incentives for DeFi builders too, because if changing your collateral requirements or risk parameters l makes you eligible for billions in additional funding, of course you’re going to build around those criteria, so we could end up with risk assessment firms influencing how entire DeFi markets are designed and the funny thing is they don’t even have to participate in them. Now take that a bit further and imagine thousands of agents relying on the same risk assessments and being able to move capital almost instantly, one downgrade could have billions trying to exit the same markets at once, sending borrowing costs through the roof and creating the exact liquidity problems the assessment was warning about lol. We could literally end up with the companies assessing risk becoming a source of systemic risk themselves. And I think that’s the part DeFi builders really need to pay attention to because although enormous amounts of financial activity is moving onchain more and more of the bargaining power and economics will move to the companies controlling the agents and the information they rely on. We’re building financial markets that anyone can access, and now some of the biggest firms in finance and tech are lining up to fight over who gets to tell the machines where everyone else’s money goes lol. It’s going to be a fun few years ahead!

It’s great to see DeFi teams ignore all the doomerism and just relentlessly ship and innovate, the guys at @3janexyz just shipped onchain credit scores with @cred_protocol and boy oh boy what they’ve been building is impressive. One of the things I’ve always found weird and funny about DeFi is how we’ve basically built an entire lending economy around capital that’s already sitting inside smart contracts, we’re so obsessed with TVL that we forget an enormous amount of tradfi operates through commitments to provide capital in the future, without that money having to sit there waiting to be used. 3Jane brought that committed capital model onchain with LCC and now they’re introducing credit scores into the equation, except the people being scored here are the ones promising to provide capital rather than the ones borrowing it. So two participants could each commit $10m of funding, but those commitments aren’t necessarily worth the same if one has a history of honouring their obligations and the other doesn’t and being able to assess and price that difference gives financial reputation actual economic value within 3janes model. I’d love to see other lending protocols adopt this because instead of competing purely on deposits you could start sourcing committed funding, with capital providers earning fees on those commitments and that means protocols can support considerable more credit capacity than their actual TVL+ open up even more design space around committed credit facilities/liquidity backstops etc. There’s still an insane amount of financial infra that we need to bring onchain, so it’s great seeing the guys at 3Jane innovating here and building this stuff around a real credit business while others are busy crying that there’s nothing left to build in DeFi lol.

Man Aave has been cooking lol, last week Sentora showed us what it looks like when curators start moving down the stack from just managing vaults into actually operating credit markets on v4 and now Hare is showing us what it looks like when sophisticated financial firms start building products on top the same infrastructure. GSR brings the balance sheet, turtle brings distribution, paxos brings the asset, hare packages the product and aave sits underneath it all providing the liquidity + credit infrastructure, which imo is where the v4 thesis starts getting really interesting because none of these guys need to build the entire financial stack themselves anymore, they can focus on what they’re really good at and plug into the same underlying system. So you can have curators moving down into market operation while asset managers, issuers and financial firms move up into product manufacturing, and instead of everyone bootstrapping isolated pools of liquidity from scratch they can all build around the same Aave balance sheet. Sentora and hare are basically approaching the same stack from opposite ends and if more businesses start doing this then you better believe aave is going to be a lot bigger than just the biggest lending protocol, I think the end game is aave beocming the wholesale balance sheet all of finance builds on. The future is bright, DeFi will win!

Ladies and gentlemen Please panic in a calm and orderly fashion.

This is really really cool, @etheconomiczone just executed the first atomic L1 L2 tx on mainnet and imo this is going to have some massive market structure implications especially around fragmentation. If you think about how eth’s scaled so far every new L2 basically creates another little economy that has to bootstrap its own apps, users and liquidity, which is why L2s have spent millions incentivising capital to bridge and ended up duplicating huge amounts of liquidity across different places. Now imagine someone wants to borrow usdc on an L2, they’d need enough usdc sitting there to fund the borrow, but if that application can access liquidity somewhere else atomically then you don’t need all that usdc sitting there in the first place. Which is why I think this is really big because moving execution somewhere new will no longer mean having to bootstrap an entire economy around that new execution environment (L2), the economic state can stay concentrated on mainnet while execution happens across loads of different places so eth can concentrate the economy while distributing the execution. That opens up a huge amount of design space for specialised rollups because a privacy L2 can just focus on being really good at privacy same for trading, AI, payments or whatever the L2 specialises in. Which is also net good for users because your going to end up with a bunch of specialised rollups all competing for order flow from the same shared economy which means execution itself becomes the product. I actually think this means we end up with way more rollups because the barrier to launching gets massively lower, and being able to tap into mainnet’s economic state while L2 teams can focus on optimising execution for their specific use case is a pretty hard proposition to reject. So if these guys keep shipping we’ll end up with one massive ethereum economy with execution happening across loads of specialised rollups instead of a thousand little economies connected by bridges, which sounds a lot more like how eth was meant to scale in the first place lol.

So everyone knows the phrase “you’ll own nothing and you’ll be happy”, I have a thesis that we actually end up with…you’ll choose nothing and you’ll be happy. Here’s what I mean, people will always own their money, assets, accounts and all the rest of it but a lot of the choices we make will disappear because when something knows what you want and can search an entire market, compare everything and actually act for you, there isn’t really a reason to keep doing all of that yourself anymore. Basically the entire internet economy was built around the assumption that people are doing the navigating, you search, open a bunch of tabs, compare options, work out who you trust and eventually choose something, and companies spend an insane amount of money trying to influence you somewhere along that journey. Agents flip that assumption on its head because the intent can still come from you without you doing most of the things that sit between deciding you want something and getting it. We’re already starting to see this with trading agents, let’s say your agent is looking at tesla, it could check the market, options, weather affecting the supply chain, satellite images of the factory or whatever else it thinks matters, put all of that together and make the trade. So technically even though you’re still the customer, you’re not really the chooser anymore. Point is we’re moving towards a machine dominated internet and so obviously there’s going to have to be some version of agent SEO where companies make it easy for machines to find and use them, it’s pretty obvious agents are going to be the main way we interact with the internet that’s why there’s so much competition between the labs (muse,dots etc) to own the agent layer. I also think finance is where this is going to be extremely important because at some point the agents are going to move from discovery to routing, in other words they won’t just help you find things but will eventually decide where the money should go. Interesting thing is DeFi is positioned really well for this because the markets themselves are software, so agents can actually see what’s happening and transact directly against the underlying contracts. We already know from aggregators how valuable it is to control orderflow because you get to influence which venues actually receive it, but agents get involved even earlier because the user doesn’t need to arrive with an order in the first place. All they have to do is tell the agent what they want and it works out how to execute that instruction and where to do it, so looks like will end up with a future where the user never actually decides which protocols to use, that choice just gets made by their agent as part of getting them the outcome they asked for. And even though this is definitely disruptive to an extent, imo it’s also one of the first major distribution shifts where DeFi starts with the advantage over TradFi instead of trying to catch up to it. So while basically the entire internet is going to be figuring out how to adapt itself for machines, DeFi is actually early lol.

Between Blast shutting down and aave taking steps towards directing more value back towards the aave token, it’s safe to say the days of useless tokens are over. End of the day if your token has no real value driver, people are eventually going to ask why it exists and the same goes for chains if they only work while you’re paying people to use it then your only renting an ecosystem. Now that the rent is due expect to see a bunch more of these useless chains shutting down. At the same time expect to see tokens with real cashflow/businesses underneath them and a mechanism for that value to accrue back to the token to massively outperform. Industry is growing up.

DeFi won this battle & the best part of it is the precedent. If you want to play with ETH issuance you need an insanely high burden of proof. “Staking is getting too popular” isn’t remotely enough to justify changing the economics of everything built on top of it.

The first credit franchise is coming to @aave v4. Crazy to see how quickly this is happening because a couple of weeks ago I posted a thesis on how v4 could let curators vertically integrate into Aave and move from simply managing vaults to operating entire credit franchises. Now @SentoraHQ is doing exactly that. I don’t think this will be an isolated case either, I expect a lot more curators to start looking at v4 and asking why they should stop at the vault layer when they can move further down the stack, operate their own credit markets and participate in more of the economics their distribution and risk expertise creates. Curators have already proven they can aggregate capital and build strategies, their next growth engine is owning more of the infrastructure those strategies sit on top of. Feels like we’re going to see a pretty big shift in what a “curator” actually is over the next couple of years. governance.aave.com/t/arfc-s…

The eth ecosystem has been absolutely cooking and this latest @ethlabs_org release is one of the best market structure upgrades I’ve seen in a while, I really love this kind of stuff because a change at the infra layer can completely change the economics one layer above it and FCR brings in some massive changes. A ton of DeFi infrastructure was basically built around the constraint that eth confirmations are slow and by significantly reducing that waiting time a lot of the assumptions around those previous constraints change too especially how liquidity has to be positioned. For example let’s say usdc borrowing on an L2 spikes to 12% while mainnet is sitting at 4%, you’d assume capital should move over and arb the difference pretty quickly, except capital can’t teleport. So you end up either waiting for it to move across or someone who already has money sitting there fronts the liquidity and settles everything afterwards which has been a moat for big market makers or firms that can afford to have millions across mainnet and L2s. Now the thing is with FCR, capital on mainnet can become usable somewhere else in seconds so the same dollar can be turned over much faster + smaller arbs become worthwhile and you don’t need as much capital sitting around doing nothing to support the same amount of flow. We also usually assume fragmented liquidity is purely a function of assets sitting on different chains or inside different pools, but some of that fragmentation is literally just time. The money exists, it just isn’t where you need it when you need it. This release helps change that, now I’m not saying that the entire ethereum ecosystem is going to magically become one giant pool, but from a financial standpoint all the different markets across the ecosystem should start pricing much more like connected venues instead of isolated pools. Like I mentioned earlier crosschain execution is basically someone saying “I already have money where you need it, I’ll pay you now and settle it later.” That someone is the solvers, now we will obviously still need them, but if capital can move much faster, warehousing inventory/capital everywhere becomes less of a moat and pricing and execution matter more which is great for users. Also because part of the fragmentation problem is literally just time, stablecoins and RWAs benefit massively because you can start separating where an asset is issued from where it actually gets used. You don’t necessarily need a bunch of liquidity sitting on every chain just because users want to transact there, issuance and redemption can stay on mainnet while the asset moves out to whichever venue needs it. The biggest thing for me though is when you look at FCR in the context of Vitalik’s world computer thesis, because then this isn’t just about liquidity moving around the ethereum ecosystem. As more external systems use ethereum as the place where state gets verified even when execution happens somewhere else, there’s naturally more reason for liquidity to sit close to that state too. FCR makes that liquidity much more useful because once something is verified on ethereum you can act on it almost immediately rather than waiting around for finality. Basically proof demand creates liquidity demand too and FCR amplifies the usefulness of that liquidity, but that’s probably a point for another day. It’s simple but kind of crazy at the same time, eth is going to be where the world proves what happened even when execution happens somewhere else and FCR means that state can be acted on very quickly giving liquidity even more of a reason to live on eth.

I think the reason the world computer paper resonated so much yesterday is because for the first time in a long time people could actually picture ethereum’s future, what the protocol’s end state may actually look like and why the world would need it in the age of AI, robotics and other frontier technologies. I mean we’re already watching AI change how we work, the jobs we do and how much of what used to need a human can now be handed off to software, which gives us a glimpse into a future where more and more human to human processes are replaced by machine to machine processes. The problem is machines can’t really operate on “trust me” in the same way humans do. Let’s say an agent’s about to process a financial transaction based on information from another machine. It needs to know the information it has is valid and the conditions are actually true, and then it needs to know the transaction really settled and everyone’s working from the same final state, which is basically the agreed record of who owns what after settlement. That’s just one transaction, now imagine that same trust problem repeated across billions of machine to machine interactions every single day. You can see where this is going. AI makes autonomous computation orders of magnitude more abundant, which means it also creates orders of magnitude more things that need to be proven and verified. At this point we all know there’s going to be billions of machines built by different labs, running different models, on different stacks and across completely different ecosystems. So the more fragmented that machine economy becomes and the more things those machines need to prove and settle with each other, the stronger the demand becomes for a credibly neutral place they can all treat as shared truth. So if you genuinely believe AI and agents are going to replace a huge amount of human coordination, I don’t know how you look at that future and not see ethereum having an absolutely gigantic role in it. It’s literally the world computer I think vitalik just saw this future before the rest of us did.

It’s getting funnier😂 they admit the post was real but their explanation is that an AI marketing tool somehow came up with 11 distributor integrations, 0%, 7%, 25% and the 80/15/5 split? Cool. Where did the AI get the numbers from?

Yoo uhm… @Morpho are you saying your business model doesn’t work?

The craziest thing about this whole Kelp/LZ lawfare situation is that it’s going to be the new normal. I don’t want to get into the legal fugazi of who’s right or wrong between Kelp and Layerzero here, but we need to understand that legal enforcement is completely normal in TradFi and as DeFi professionalises I think we’re going to see a lot more of it. DeFi tries to replace financial intermediaries with code, but code replacing execution and settlement doesn’t mean disputes around negligence, disclosure, responsibility or who eats the loss when something goes horribly wrong suddenly disappear. In TradFi if you rely on critical infrastructure and it fails there are contracts, warranties, indemnities, insurance and the legal system as an avenue of recourse. And I think as more institutions come onchain a lot of that will start making its way into DeFi too, which probably changes how we think about due diligence on critical infrastructure and “we have 5 audits bro” isn’t going to cut it anymore. We’re going to have to start asking questions like who is actually standing behind this, what have they represented, what happens if their infrastructure fails, what insurance do they carry and who is liable for the loss. Which is also going to significantly change the competitive landscape because being able to actually stand behind your infrastructure financially and legally starts becoming part of the security model itself. Smart contracts aren’t going anywhere, but neither are actual contracts.

This is a really strange argument because the argument basically defeats itself. You start by explaining that vaults exist because users can’t realistically track and manage thousands of markets themselves, so the vault abstracts that complexity and someone else allocates the capital for them. I don’t think anybody could have thought of a better definition for delegated portfolio management aka asset management lol. Also a timelock doesn’t magically turn the curator’s decision into the depositor’s decision, it just gives the depositor notice of any changes, your own docs literally say the Curator configures “liquidity allocation rules”, is “abstracting risk curation decisions away from depositors” and makes key decisions about “how capital is allocated.” So even if the code constrains the manager, it doesn’t remove the manager. You also failed to mention what SEC commissioner pierce actually says and thats vaults fall on a spectrum between “programmatic allocations determined solely by immutable smart contracts” and “allocations at the sole discretion of another person or group of persons.” She also explicitly points to selecting yield generating activities and reallocating assets as examples of managing a vault and says managing vaults can raise investment adviser issues. You’re basically trying to make the asset manager disappear by changing the test from “who is making the investment decisions?” to “can they steal the money and can I withdraw?” those are completely different things. Even your “implicit approval” argument is backwards, if I delegate allocation to you then you announce a change and I don’t withdraw during the timelock, my failure to leave hasn’t somehow transformed your investment decision into mine. In fact that requires me to continuously monitor the manager which is the exact complexity vaults supposedly exists to abstract away in the first place. Vaults can absolutely be noncustodial but noncustodial settlement is not non discretionary allocation, no matter how you try to spin this an asset manager is an asset manager even if you give them cute sounding names like “curator” or “allocator”.

Gearbox is a great example of why you shouldn’t confuse lack of attention with lack of progress, the market hasn’t exactly treated them kindly, but while everyone moved onto the next narrative they just kept grafting on the same core architecture. This RWA leverage solution they’ve just released is where that strategy is really going to pay off imo. I’ve spent a bunch of time looking into the different approaches being built here but what @GearboxProtocol has done is by far the most elegant design I’ve seen. It’s not just better looping ux, looping naturally assumes the underlying asset is liquid and everything settles atomically so you borrow, buy the asset, deposit it again and repeat, which is great for crypto native assets but doesn’t really work for RWAs where you have different redemption timelines, transfer restrictions, KYC requirements and in a lot of cases barely any secondary liquidity. So gearbox gets around all these issues by using the credit account to borrow the full amount upfront and subscribe directly with the issuer. Instead of needing 5 or 10 or however many separate loops and waiting through the settlement process each time, you can create the entire leveraged position in one go and redeem the entire thing in one go on the way out. What I really like about this is that leverage no longer needs to depend on secondary market liquidity to the same extent, because it was always pretty silly to expect an issuer to bootstrap $50m or $100m of dex liquidity before people can take levered positions on the asset. With this model the position can scale against available credit instead because the credit account is interacting directly with the issuer, which means the lending side can support assets that would have been impossible or just really inefficient to support through the normal loop model. Even the asset specific stuff like KYC, transfer restrictions, redemption and specialised liquidation logic can sit inside gearbox’s infra rather than the lending protocol having to figure all of that out. And because gearbox is separating the asset specific execution and risk machinery from the funding layer, I think the bottleneck for lending protocols changes quite a bit too. Because the problem is no longer whether an asset has enough secondary liquidity to be listed and what really starts to matter is whether somebody can actually originate good borrow demand against the liquidity sitting there. If an issuer can bring the asset, gearbox can handle the market specific plumbing and a lending protocol can provide the funding, which is just a more efficient way of creating leverage around assets that never really fit the normal money market model in the first place. These guys have done a great job of addressing the actual market specific pain points and creating a structure where everybody wins, RWA issuers get leveraged distribution without needing deep secondary markets first, stablecoins get a new structural source of borrow demand and lending protocols get access to a much wider set of credit opportunities. What I’m really excited to see now is how much leverage this can support in practice, because if you can get a decent amount of leverage on assets that were basically unleveragable before then you open up a pretty massive new market. Really excited to see how this plays out and wishing the chads at gearbox all the best with this launch.

I’ve been saying for a long time that v4 is a huge opportunity for curators because it can turn them from vault managers into actual lending franchises, so it’s great to see Stani validate the thesis and clear up some myths around v4. I really think the operating leverage v4 gives curators and integrators is pretty insane. Curator growth is tied to AUM but with v4 they can build from the demand side instead, find the borrowers, originate the credit, create the spoke and draw liquidity from the hub. If you’re good at that, the same funding base can support an expanding book of completely different credit markets. Which changes the economics of the curator business because your growth is no longer capped by how much capital you can attract into the next vault and is now driven by how much good credit demand you can originate. So at scale I really think we could have curators running multi billion dollar specialist lenders on top of Aave without ever needing to build the underlying liquidity network themselves. That’s a massive upgrade to the model and a much bigger business than charging 10-20bps to decide where someone’s USDC goes.

I too would write a love letter about partnering with competitors if I’d paid Robinhood $100m + for the privilege lol. The funniest part is coinbase then turning around and bringing aave in as a credit layer for tokenised stocks. The issue here is the conflating of distribution with a network effect that somehow accrues to morpho. Coinbase using morpho doesn’t make robinhood more dependent on morpho and vice versa, If anything, every big distributor you add has more incentive to make sure the credit layer stays competitive and replaceable. And I don’t think “we power your biggest competitor too” after paying them an insane amount of money is quite the flex to coinbase that you think it is. The base + Aave stuff on tokenised stocks is basically showing you the model in real time. Coinbase and RH own the users and distribution, they can shop the credit layer around product by product and make the protocols compete underneath them which is exactly what RH did and why you ended up massively overpaying. It’s basically the supermarket model they own the shelf space while you’re fighting to be stocked which means: You’re not the network. You’re a vendor to the network.