AMMs are a constrained primitive. They define a single, rigid trade execution path: deposit liquidity into a public pool and accept the resulting price. This design creates systemic inefficiencies like high slippage and MEV extraction that more expressive systems solve.
history-of-money-and-the-crypto-thesis
Blog
Why Automated Market Makers Are Just the First Primitive
AMMs like Uniswap revolutionized trading with a simple, stateless function. But they are just the first step. The next wave is intent-based, stateful solvers that optimize for user outcomes, not just execution. This is the evolution from programmable liquidity to programmable trade settlement.
introduction
THE PRIMITIVE
Introduction
Automated Market Makers are a foundational but limited primitive, setting the stage for more expressive, intent-driven architectures.
The next evolution is intent-based architectures. Protocols like UniswapX and CowSwap separate user preference (the 'intent') from execution, allowing for off-chain order flow aggregation, MEV protection, and multi-venue routing that AMMs cannot natively provide.
This shift mirrors web2's API evolution. Just as monolithic databases gave way to specialized microservices, monolithic AMM pools will fragment into modular components for solving, routing, and settlement, as seen in the Across bridge and 1inch Fusion models.
thesis-statement
THE EVOLUTION
The Core Argument: From Stateless Functions to Stateful Solvers
AMMs are a foundational but limited primitive, and the next evolution is towards stateful, capital-efficient solvers.
AMMs are stateless functions. They execute price discovery via a deterministic formula (x*y=k) with no memory of past trades or external data, making them simple but capital-inefficient.
Solvers are stateful agents. Protocols like UniswapX and CowSwap use off-chain solvers that consider order flow, liquidity across venues, and MEV to find optimal execution paths.
The shift is from passive to active liquidity. An AMM is a passive pool; a solver actively routes through Curve, Balancer, and bridges like Across to minimize price impact.
Evidence: UniswapX now routes over 50% of its volume via solver networks, demonstrating superior fill rates and cost savings versus the v3 AMM alone.
key-trends
FROM AMMS TO INTENTS
Key Trends: The Rise of Intent-Based Architectures
Automated Market Makers solved price discovery but left a fragmented, inefficient user experience. Intent-based systems abstract this complexity by letting users declare what they want, not how to get it.
01
The Problem: AMMs Are Execution-Layer Primitives
AMMs like Uniswap V3 are brilliant for liquidity but force users into a single venue. They are low-level building blocks, not user-centric interfaces.\n- Inefficient Routing: Users get the pool's price, not the best cross-DEX price.\n- MEV Vulnerability: Public mempools expose simple swaps to front-running.\n- Fragmented UX: Requires manual aggregation across chains and protocols.
~30%
Potential Slippage
$100M+
Annual MEV Loss
02
The Solution: Declarative Intents & Solver Networks
Projects like UniswapX, CowSwap, and Across let users submit signed intent messages (e.g., 'Swap X for Y at best rate'). A competitive network of solvers (like 1inch Fusion or Flashbots SUAVE) fulfills it off-chain.\n- Price Optimization: Solvers compete, guaranteeing the best execution across all liquidity sources.\n- MEV Resistance: Intents are private orders, not public transactions.\n- Cross-Chain Native: Solvers use bridges like LayerZero and Wormhole as mere components.
10x+
Liquidity Sourced
-99%
Failed Tx Rate
03
The Endgame: Composable Intents as the New Standard
Intents evolve from simple swaps to complex, conditional statements—'Bridge to Arbitrum if gas < $5, then provide liquidity and stake the LP token.' This turns Ethereum into a settlement layer for verified outcomes.\n- Composability: Intents become inputs for other intents, creating meta-protocols.\n- Specialized Solvers: Emergence of solvers for derivatives, lending, and identity.\n- User Sovereignty: Retain asset custody until fulfillment, reducing counterparty risk.
$10B+
Future TVL
~500ms
Quote Latency
PRIMITIVE EVOLUTION
AMM vs. Intent-Based Architectures: A Feature Matrix
A first-principles comparison of the dominant on-chain liquidity primitive (AMM) against emerging intent-based systems like UniswapX, CowSwap, and Across.
| Core Feature / Metric | Classic AMM (e.g., Uniswap V2/V3) | Intent-Based Architecture (e.g., UniswapX, CowSwap) | Hybrid Solver Network (e.g., Across) |
|---|---|---|---|
Execution Model | Passive, Pre-Defined Curve | Declarative, User-Specified Outcome | Auction to Professional Solvers |
Price Discovery | On-Chain Bonding Curve | Off-Chain RFQ & Competition | Competitive Solver Bids |
Maximal Extractable Value (MEV) Exposure | High (Front-running, Sandwiching) | None (Intents are private until execution) | Low (Auction mechanism captures value for user) |
Gas Cost for User | User pays for all on-chain tx gas | User pays 0 gas (Sponsored by filler) | User pays 0 gas (Sponsored by solver) |
Typical Price Improvement vs. AMM Quote | 0% (Baseline) |
|
|
Cross-Chain Capability | False (Requires bridging assets first) | True (Native via fillers like Across, Socket) | True (Core design purpose) |
Liquidity Source | Single Pool's Capital | Any On/Off-Chain Venue (DEXs, OTC, Private MMs) | Canonical Bridges + AMMs |
Settlement Finality | 1 Ethereum Block (~12 sec) | 1-5 mins (Batch auction window) | 10-20 mins (Optimistic challenge period) |
deep-dive
THE AMM EVOLUTION
Deep Dive: How Solvers Rewire DeFi Economics
Automated Market Makers are a foundational but incomplete primitive, now being superseded by intent-based architectures that externalize execution logic.
AMMs are a bottleneck. They embed execution logic directly into the contract, forcing users to accept a single, predetermined price discovery path. This creates persistent inefficiencies like MEV extraction and liquidity fragmentation across pools like Uniswap V3.
Solvers externalize execution. Protocols like CowSwap and UniswapX shift logic to a competitive off-chain network. Users submit intent-based orders stating what they want, not how to achieve it. Solvers compete to find optimal routes across DEXs, bridges like Across, and private markets.
This inverts the fee model. AMMs collect fees from passive LPs for providing a service (liquidity). Solvers earn fees for providing a result (best execution), creating a performance-based market where efficiency is directly monetized.
Evidence: CowSwap's CoW Protocol now routes over 68% of its volume via its solver network, bypassing on-chain AMM pools entirely and demonstrating the economic viability of this new architecture.
protocol-spotlight
BEYOND THE AMM
Protocol Spotlight: The Vanguard of Intent
Automated Market Makers (AMMs) like Uniswap and Curve solved liquidity bootstrapping but introduced rigid, costly execution. The next wave of primitives uses intents to abstract complexity and optimize for user outcomes.
01
The Problem: AMMs Are Dumb Pipes
AMMs are passive, reactive liquidity pools. They execute trades at the worst possible price within a block, creating a ~$1B+/year MEV opportunity for searchers. Users pay for inefficiency.
- Price Impact & Slippage: Large trades move the market against the trader.
- Fragmented Liquidity: Capital is siloed, failing to find the best price across venues.
- Atomic Execution Only: No ability to route through private order flow or off-chain solvers.
$1B+
Annual MEV
>30%
Slippage (Large Trades)
02
The Solution: UniswapX & CowSwap
These intent-based protocols let users declare what they want, not how to achieve it. Off-chain solvers compete to fulfill the intent optimally.
- MEV Protection: Solvers internalize value, returning it as better prices.
- Gasless Signatures: Users sign intents, paying fees in output tokens.
- Cross-Chain Native: UniswapX aggregates liquidity across Ethereum, Arbitrum, Polygon, and Base via Across Protocol.
~$10B+
Volume Processed
-99%
Gas for User
03
The Architecture: Solver Networks
The magic is in the solver competition. Entities like CowSwap's decentralized network and UniswapX's permissionless solvers bid to fill orders.
- Batch Auctions: Orders are settled in discrete time intervals, reducing frontrunning.
- Optimal Routing: Solvers can tap into on-chain AMMs, private inventory, and OTC desks.
- Credible Commitments: Solvers post bonds, ensuring they execute winning bids.
~500ms
Auction Resolution
100+
Active Solvers
04
The Future: Intents as a Universal Abstraction
Intent-based design is spreading beyond swaps to bridging, lending, and portfolio management. Across Protocol uses intents for cross-chain transfers, while Anoma envisions a fully intent-centric architecture.
- Composability: A single intent can trigger a multi-step, cross-protocol transaction.
- User Sovereignty: Users specify constraints (e.g., "swap only if price > X").
- Infrastructure Shift: The stack moves from RPC endpoints to intent propagation networks and solver markets.
10x
Complexity Abstracted
LayerZero
Competing Standard
counter-argument
THE ECONOMICS
Counter-Argument: Are Solvers Just Expensive Aggregators?
Solvers are not just aggregators; they are a new primitive that optimizes for finality and composability, not just price.
Solvers optimize for finality, not price. Traditional aggregators like 1inch find the best price across existing liquidity pools. Solvers like those on CowSwap or UniswapX compete to guarantee the best settled outcome, which includes MEV extraction, gas optimization, and cross-chain execution via Across or LayerZero.
Their cost is a function of complexity. A simple swap is cheap. A solver's fee for a multi-leg, cross-chain intent that bundles a swap, a bridge, and a deposit into a yield protocol is the cost of orchestrating atomic composability. This is a different service.
The market validates the model. The volume settled via intent-based systems like UniswapX and CowSwap proves users pay for guaranteed execution. They avoid failed transactions and sandwich attacks, which are implicit costs in the AMM model.
Evidence: UniswapX processed over $10B in volume in its first year, with solver competition driving down fees for complex trades that would fail on vanilla AMMs.
takeaways
BEYOND AMMs
Key Takeaways for Builders and Investors
Automated Market Makers like Uniswap and Curve solved price discovery, but they are just the first primitive. The next wave of DeFi innovation is about solving for execution, composability, and user intent.
01
The Problem: AMMs Are Dumb Pools of Capital
AMMs passively wait for arbitrageurs to correct prices, creating a permanent loss vs. rebalancing cost dilemma for LPs. They are execution-agnostic, unable to source liquidity from better venues.
- Inefficient Capital: Billions in TVL sit idle, earning minimal fees.
- Slippage & MEV: Users pay for poor execution via front-running and sandwich attacks.
$10B+
Idle TVL
>90%
Passive LPs
02
The Solution: Intent-Based Architectures (UniswapX, CowSwap)
Shift from specifying how (a swap) to declaring what (a desired outcome). Users submit signed intents; a network of solvers competes to fulfill them optimally.
- Better Execution: Solvers route across AMMs, RFQ systems, and private inventory.
- MEV Protection: Intents are settled in batches, neutralizing front-running.
- Gasless UX: Users sign once, solvers pay gas.
~30%
Better Prices
0 Gas
For User
03
The Problem: Liquidity is Siloed and Static
AMM liquidity is trapped in individual pools and chains. Cross-chain swaps require a daisy-chain of vulnerable bridges and AMM hops, creating a terrible user experience and systemic risk.
- Fragmented TVL: Capital efficiency plummets across 50+ L2s.
- Bridge Risk: Over $2B+ has been stolen from bridges, the #1 attack vector.
50+
Liquidity Silos
$2B+
Bridge Hacks
04
The Solution: Universal Liquidity Layers (Across, LayerZero, Chainlink CCIP)
Abstract liquidity into a network-layer resource that can be programmatically deployed anywhere. These systems use optimistic verification, atomic commits, or decentralized oracle networks to move value.
- Capital Efficiency: One liquidity pool can serve dozens of chains.
- Unified UX: Native cross-chain swaps feel like a single transaction.
- Security: Move away from monolithic, hackable bridge contracts.
10x
Efficiency Gain
<2 mins
Settlement
05
The Problem: On-Chain Activity is Opaque and Unstructured
Transactions are low-level bytecode calls. Wallets and apps cannot reason about user goals, making advanced features like transaction batching, fee sponsorship, and secure session keys impossible.
- Wallet Limitations: Every action requires a new signature and gas payment.
- No Abstraction: Users must understand gas tokens, approvals, and slippage.
5+
Clicks per Swap
High
Cognitive Load
06
The Solution: Account Abstraction & Programmable Transactions
Make user accounts programmable smart contracts (ERC-4337). This allows for social recovery, batched operations, and sponsored gas fees paid in any token.
- Session Keys: Grant limited permissions to dApps for a seamless experience.
- Transaction Bundling: Execute a complex DeFi strategy in one click.
- Mass Adoption: Removes the seed phrase and gas token barriers.
1-Click
Complex Actions
Any Token
Pay Gas
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall