# On DeFi, DeFi*, and DeFAI Source: https://substack.com/home/post/p-154325327 Kiwi News: https://news.kiwistand.com/stories/On-DeFi-DeFi-and-DeFAI?index=0x6964e17b6fe3b28fe7cc68eaa1dbeaa868a63b25104ccb6db5d2cb4717a138213a9d26ae Submitted by misha on 2026-01-12 UTC, 3 upvotes, 0 comments ## Summary The article argues that DeFi is having a quiet resurgence and that its enduring product-market fit depends on whether protocols keep cypherpunk properties or compromise them for UX. It distinguishes "DeFi" (or DeFi Classic), which prioritizes decentralization and censorship resistance, from "DeFi*", which accepts compromises for UX, best execution and competitiveness with offchain alternatives. The author's stated intuition is that DeFi represents an opportunity in the trillions, while DeFi* may represent one in the quadrillions. The article also cites newer verticals such as DePIN, stablecoin commerce (NoFi), and DeFAI, with Helium's $HNT and $MOBILE tokens as an example, though it notes that some narratives are running ahead of fundamentals. ## Article DeFi all the way down? One could argue that DeFi is having a moment, even if a quiet one. While the original smart contract use case with real traction has fallen somewhat out of fashion over the past 18 months as a new bull run demanded new narratives (AI right now thoroughly dominating the conversation), crypto's humble and arguably sole source of true product market fit keeps chugging away. With restaking transforming the yield and productive asset landscape over the last year, the launch and success of Hyperliquid, Ethena and up and coming protocols like Morpho Blue, the success of Solana's monolithic DeFi ecosystem and the much-anticipated launch of Uniswap v4, Berachain and Monad, DeFi’s steady return to vogue is is promising to supercharge crypto's main burning ember of PMF. DeFi itself is no longer the leading crypto narrative and hasn’t been for a while (AI has that current crown - more on that later), but represents the bedrock of crypto PMF and as pure infrastructure plays lose their luster outside of an elect few, those looking for the so-called “app play” are starting to refocus more on the basics of crypto’s oldest and arguably most enduring use case producing real protocol revenues for infrastructure and apps alike. While we are seeing DeFi itself mature in scalability, expressivity and UX, the potential fatal flaw for DeFi is of course this: to the extent that cryptoasset flows are determined by speculation on asset prices, DeFi cashflows and protocol TAM will be fundamentally tied to those market cycles and the volatility they bring, unless countercyclical use cases such as products aimed at everyday consumers non-speculative financial and commercial lives are cultivated. The degen casino will undoubtedly provide fodder for the indefinite future for DeFi protocols and apps, but even the most efficient possible system for moving tokens around requires that those tokens be valuable in some way, be that a speculative premium, or more enduring PMF-driven cashflows. This is precisely what we are starting to see happen as DeFi begins to “go everywhere” with things like DePIN, NoFi (my broader term for stablecoin commerce), TravelFi, GameFi and now “DeFAI,” even as Tarun points out in the case of DeSci, narratives in these areas are running well ahead of fundamentals to put it charitably. One expects there to be a typical over-hype, over-speculation, disappointment and then ultimately plateau of productivity in most of these areas. And of course, the level of decentralization and cypherpunk alignment exists on some kind of varied spectrum for all of these different vertical specific use cases that all represent some kind of “X-Fi.” As I was writing this post, Ryan Watkins put it pretty succinctly so I’ll save myself some time by just quoting him. There is genuine innovation and discovery going on in the application of DeFi primitives like BFT consensus, smart contracts, DEX’s, stablecoins, NFTs and lending protocols to a diaspora of vertical specific use cases where the tech often plays an enabling or sidecar role inherent to the use case (usually where money changes hands and entities need to interact and transact with one another). And while one could justifiably take the radical pragmatist view that Toly expresses about there really only being a few smart contracts that have yet shown they need to exist, I’m confident that that number is going to go up. Stablecoins and their immediate use cases like remittances, payments and inflation resistance (NoFi in my parlance) are themselves built on top of DeFi primitives like ERC-20s, multisig wallets, AMMs and lending CDPs and offer the potential for basic banking, payment and investing services to be made more efficient and accessible. DePIN networks like Helium have a complex budding economy powered by the $HNT and $MOBILE tokens as a way of delivering cheaper, user-owned internet access. We now have Kaito self-branding their mindshare and social clout tokenization project Yaps as "InfoFi," and projects like Camino building tooling for the travel industry which leverages permissionless token standards, NFTs and and stablecoin payments. The Cosmos universe is on the bleeding edge of self-sovereign community finance innovation with projects like Anoma and Cycles, where the economic unit of the “firm” is not taken for granted. And of course, we are seeing the wave of AI agent experimentation (and speculative mania), and regardless of the short term bubble, it is hard to imagine every possible agent x crypto use case stone being unturned and some enduring PMF not being found, be with agents as simply a new interface for users to interact with onchain, or the much more avant-garde autonomous sentient meme vein eventually striking gold. Perhaps more philosophically, even almost entirely non-financial use cases like the core Farcaster smart contracts or ENS and EAS rely on the cryptoeconomic hardness of blockchains as tamper-proof replicated state machines. Decentralized computation and consensus is inherently DeFi. An open, global financial system underpins all of this, from the lowest level economic security, all the way to the application layer and how tokens are used in apps and games. It is finance in the way that SaaS is finance as and where it touches the financial and identity rails of the real world - and like the concept of "embedded finance" in web2, crypto will deploy "embedded DeFi" into all aspects of the economy if it is successful. A tale of two DeFis As design decisions and competition complete their intricate dance with one another while the market develops and projects explore evolutionary niches like those mentioned above, I would submit that we are starting to see what can be usefully conceived of as Two DeFis emerging: DeFi, or DeFi Classic, which prioritizes the decentralization and other cypherpunk ideals, and posits a dimension of product-market fit that relies on those ideals and system properties, and DeFi*, DeFi which utilizes public blockspace, but which makes compromises on one or more of the more extreme cypherpunk ideals in the name of UX, best execution and competitiveness with offchain alternatives. My intuition is that DeFi represents an opportunity in the trillions, but DeFi* may represent one in the quadrillions, as the latter represents a more plausible medium term technological migration that provides strict benefits to TradFi and the broader economy as it currently exists, but while making serious compromises on what the more pure DeFi protocols are ready to make. This dichotomy is really more of a dialectic or spectrum and I am careful not to draw hard and fast lines necessarily, but I think it's worth considering how these worlds are starting to diverge, and the potential challenges and opportunities that presents for crypto's only real golden goose (for now). Without adding a moral lens to it intentionally, the analogy of Augustine’s City of God versus the City of Man comes to mind, where elements of both DeFi and DeFi* are present simultaneously in varying quantities across some distribution of protocols and applications. DeFi and the true believers Before there was much enterprise interest outside of "DLT" and innovation department driven "supply chain experiments," the smart contract application era of crypto inaugurated by Ethereum was rife with utopian ideals and grand societal visions for immutable, eternal protocols or hyperstructures running on this newly invented substrate of cryptoeconomic hardness. Finance was the most natural proximate opportunity to apply crypto's newfound expressivity, both because of crypto's financial nativity and native digital asset architecture, but also because much of the early ideological motivation crypto builders brought to bear was centered around a critique of contemporary finance, dominated by intermediaries, opaque and corrupt institutions, and unnecessary friction to innovate on financial services in the open. The merits of that initial formative critique of the financial system notwithstanding (I'll save whether India's UPI or Brazil's Pix represent a sufficient upgrade technically and socioeconomically, or whether unfiltered access to onchain gambling is really making the world a better place for another post), the cyperpunk ideals and their application to finance was the initial seed that eventually bloomed with the first DeFi summer and Compound, Uniswap and Aave provocatively making the case that crypto could indeed host a completely sovereign, independent financial system. Among these principles (they're almost like some kind of crypto creed at this point, I lay the big 6 out in this post in a bit more detail) were of course decentralization (nobody was running these financial products in particular), censorship resistance (anyone with a wallet could use them without being blocked, so long as they could submit a tx onchain), credible neutrality (the protocols were designed in a way to not advantage insiders or players with special market access), radical composability (smart contracts were essentially applications with totally open APIs that could be combined in arbitrary ways), and noncustodiality (your keys, your coins). One might also tack on privacy and pseudonymity as at least core values held in certain aspects of crypto from the very beginning, albeit with the vast majority of crypto value still passing through completely transparent and un-obfuscated chains, and “multilateral atomic settlement” as Ethan Buchman puts it as a unique property of the clockwork of crypto that could be posited alongside these others. While these properties and core values of cryptosystems seemed then and now somewhat ivory tower in their essential character, they played a real role in the first wave of DeFi protocol and capital formation, from the regulatory arbitraged capital raises that the ICO era made possible, to allowing for tokens to be created and traded instantly on an AMM, or airdropped to the users of protocols, nothing about DeFi's beginning would have happened without these cypherpunk aligned affordances. Great, so far so good, crypto's idealistic and radical properties are starting to enable new use cases which strictly require those properties, exactly the kind of thing you'd want to see as a technology tries to achieve PMF in a certain vertical. There is a vibrant, albeit potentially less noisy, side of DeFi and all its derivatives that is chugging along and building public goods and being bearers of Vitalik’s Good News. DeFi* and the better mousetrap But of course, every origin story has its Fall, and as the crypto movement began to give way to the crypto industry, we have begun to see defections from the most stringent of cypherpunk ideals in the name of UX, scalability and adoption. We've seen high-performance blockchains which make fundamentally different tradeoffs around home staking, validator decentralization and stake decentralization, and seen them able to draw real usage and adoption even among those crypto natives who purport to hold these values. The trading and MEV supply chain is rife with centralized and centralizing compromises, especially where UX and trade execution matter most, as they do in TradFi. And of course most users access onchain applications through the help of centralized (sometimes more-or-less self-custodial) wallet applications, centralized RPC providers and frontend applications which are forced by jurisdiction to act discretionally in serving users and to follow the law. Everything exists on a spectrum, and without digging up every debate in the landscape of blockchain scaling and DeFi mechanism design, making hard categorizations (L2Beat is best prepared to do that) or naming and shaming any particular projects or designs, I'll just state it as the fact that we all know: we have made compromises to solve for other product dimensions than crypto's most core and unique properties and values, and that is actually probably okay. It's okay because on that spectrum, you have space for the current economy to utilize some of the same infrastructure that we are using to bootstrap a new economy. Tradfi, web2 and the broader existing economy are beginning to get sucked into crypto's technological vortex, not just its speculative one. Setting aside the new class of crypto-rich HNWIs that wealth managers are trying to reach, and even setting aside institutional crypto trading and custody solutions, which are really just extensions of the tradfi economy servicing a profitable niche, we have real interest from tradfi in experimenting with asset issuance, trading, rehypothetcation and all manner of financial product design using the affordances of public blockchains. The compromises made on things like centralized sequencing, sanctions and KYC/AML compliance, self-custody and otherwise will enable these institutions and their customers to better reap the benefits of onchain (and pay for gas like the rest of us) while not committing to an extralegal political cypherpunk movement. But what are we doing here? Does the DeFi-DeFi* distinction matter? But wait a minute, if it isn't for those things which makes crypto unique, if those things are compromised on, what is TradFi even doing here? Why are people taking UX shortcuts to build crypto products for the “real economy” to begin with? What is interesting about crypto when you remove one or more of the sacred properties? Colludingnode would convincingly argue absolutely nothing, and essentially point to regulatory arbitrage as the only logical conclusion of the “why bother with blockchain to begin with if you’re going to not go full cypherpunk.” And I am somewhat sympathetic to this argument: all of the use cases I’ve described above, and most others that one could come up with, are still in their infancy in terms of adoption. While Hivemapper announced today that they’ve now mapped 30% of the earth with their user-owned DePIN mapping network, we are still firmly in the planting season, not the sewing season, on anything else but DeFi (and it’s inherent casino cyclicality). The null hypothesis here is probably that crypto really exists to fund speculation and a number of regulatory arbitrage style use cases, whether that be in capital formation, remittances and payments, forex or otherwise. But indulging the optimistic case, it’s a well trodden path for industries that are being reimagined by new tech to run ahead of the letter if not the spirit of the law, particularly where there is a vacuum of regulatory clarity, in hopes of building enough value that under principles of capitalist legitimation through the demand of the market, that “asking for forgiveness, not permission” principle that Uber so aggressively exemplified seams reasonable enough to grant as a part of what is going on. When you think about it, it’s remarkable that Larry Fink himself, the CEO of one of the most powerful organizations on earth, touts Bitcoin’s extralegal sovereignty from governments and institutions as part of its attractiveness as an investment. The establishment financial sector and its regulators may have been relatively hostile to crypto up until recently, but the actually politically revolutionary aspects of crypto are at the same time being evangelized by some of the establishment’s most blue blooded emissaries. Capital controls in in emerging markets, where unfettered dollar access and monetary policy leakage (the flipside of the “inflation resistance” value proposition) read more like threat to global central banks than an opportunity, are likely to be among the stickiest points of regulatory pushback on the very premises of crypto’s value proposition. But we live in a multipolar legal and commercial world system, with jurisdictional competition for investment, and crypto has enough of a hold in enough places around the world as an industry and, to a lesser extent, a constituency to very likely be able to end up with some kind of middle ground set of compromises with the traditional order. This is, after all, the very global order which produces and legitmizes TradFi tax havens that the elite who write the rules flock to. Less cynically, and perhaps if you take the “move fast and break things” advantages that indestructible (and sometimes incomprehensible) crypto affords the mercurial builders for granted, radical composability and the speed of innovation and experimentation it allows for is the quantity-that-becomes-quality aspect of crypto’s unique edge over centralized servers. The public standards and interfaces that things like the EVM, ERC-20 tokens or the Uniswap v2 swap router provide as building blocks give us a glimpse as to what can be achieved through permissionless composability. Panoptic has built an options protocol on top of Uniswap v3 LP positions, for example, and Uniswap v4’s constrained-but-open platform protocol design creates a form factor for experimentation around order types while maintaining discoverability by solvers/sources of liquidity and flow. This kind of experimentation, especially as and where it would otherwise require intermediation and several different parties to coordinate affirmatively, can be done more quickly in the crypto wild west than it can be done in TradFi or web2 itself. Sure, completely open API standards and transparent pay-per-use service agreements could develop in finance, web2 marketplaces or web2 gaming, but they are really hard to coordinate as a federation of centralized entities. Blockchains, smart contracts and standard logical interfaces that can be accessed by anyone make this kind of combinatorial experimentation easier. Whether that experimentation will yield something both profitable and ultimately legal I’ll leave as a thought experiment and open question for the reader. Crypto and blockchains also represent a truly global, 24/7 market, something which capital has abstractly always been moving towards and shaping the global system around, and now which it finally has in a fully unmediated form. Firms from two geopolitically complicated home nations might be able to transact over a shared neutral substrate without fear of being cut off by one or more of those national governments (or a third party government). Less radically, entire classes of intermediaries required for payments, purchases or the transfer of securities or other assets to reach the other side of the world can potentially be eliminated by smart contracts and the assurances that the underlying blockspace give to the parties relying on them. T-N settlement times can be reduced down to blockchain finality time and allow for capital formation, commerce and trade to exist at the absolute cutting edge of what their jurisdictions will legally allow. Capital and commerce, being that irresisistible solvent that melts away boundaries, will be able to use crypto to transact borderlessly. People, businesses and communities have access to global, always-on uncensorable computation, upon which varying levels of compliant computational and economic activity can take place. Much of the above amounts to crypto, once UX is solved (and I argue it’s basically 90% on the way to being solved), being useful inasmuch as it reduces friction versus centralized counterparts. Sending stablecoins, given a sufficient install base and all of the privacy and fiat onramping caveats, will be strictly faster, cheaper and easier than sending money through the traditional system for almost any use case. One cannot really imagine a web of AI agents autonomously transacting with one another, acquiring resources like compute and storage, and interacting with the real world through traditional bank accounts. As the crypto rails themselves are able to disappear an the actual use cases, the bull case for DeFi and DeFi* protocols and applications ultimately comes down to actual reduced economic friction in the real world, as an abstract application of all of or some of its unique affordances to real world problems. Crypto in general, and DeFi in particular (but I repeat myself), will live or die on whether it can actually realize this reduction of friction in real product-market fit that it uniquely enables.