We are told that Layer-2 solutions have finally solved Ethereum’s gas fee problem. Optimism and Arbitrum are processing transactions for pennies. Base is onboarding millions. The narrative is that scaling is done, that Ethereum is now a multi-chain universe with infinite capacity. But what if this is the same illusion that gripped oil executives in West Texas six months ago?
In early 2024, new pipelines finally broke the massive gas glut in the Permian Basin. For months, natural gas was practically worthless at the wellhead — trading at negative prices because there was simply no way to move it to consumers. Then the pipelines opened. Prices stabilized. Producers breathed a sigh of relief. Local economies cheered. And then, almost immediately, the drilling plans came back. The same companies that had been choked by oversupply started filing permits to drill more wells. Because that’s what a bull market does: it tricks you into thinking the bottleneck was the only problem.
The same pattern is playing out right now on Ethereum’s Layer-2 ecosystem. We have built the pipelines — Arbitrum Nitro, OP Stack, zkSync Era, Base. These rollups are the new interstate highways for transactions. They have absorbed the congestion that made mainnet unusable. Gas fees on Ethereum have dropped to under 5 gwei. Users are flooding into L2s. But here is the uncomfortable truth that no one wants to say out loud: the relief is real, but it is temporary. The pipelines are easing the glut, but the drilling plans — the wave of new applications, token launches, and speculative activity that a bull market enables — are already in motion.
Let’s look at the data. Total value locked on Ethereum L2s has grown from $5 billion in January to over $15 billion by June. Daily active addresses on Base alone have surpassed 1 million. The engineers have done their job. Throughput is up, costs are down. But here is what my experience as a protocol PM has taught me: capacity does not equal sustainable equilibrium. Every time we solve a throughput bottleneck, we invite a new wave of demand that was previously suppressed. In the Permian, the pipeline opened capacity to 2 billion cubic feet per day. Within three months, producers had filed permits to add 400 million cubic feet of new supply. The cycle resets.
Now look at the Ethereum ecosystem. The Dencun upgrade in March 2024 slashed L2 data costs by 90%. Blob space was cheap. The immediate effect was a surge in L2 activity. But here is the hidden signal: the number of new token contracts deployed on L2s has increased by 300% since March. Every one of those tokens will bring its own liquidity, its own user base, its own transaction demand. The mempool is expanding faster than the pipelines can scale. The development teams are already planning the next wave of upgrades — Proto-Danksharding Phase 2, peer-reviewed rollup improvements — but they are running a race against the very demand they are enabling.
This is the paradox of decentralized scaling. Decentralization is a verb, not a noun. It is a continuous act of balancing capacity and demand. The moment you think you have solved scaling, you have already created the conditions for the next bottleneck. We saw it with Ethereum mainnet in 2021. We saw it with Solana in 2022. We are seeing it now with L2s in 2024. The infrastructure improves, and then the applications flood in to fill the vacuum. It is not a bug. It is the nature of permissionless systems.
The contrarian angle here is uncomfortable for the bull market narrative. Right now, everyone is euphoric that gas fees are low. Developers are shipping products. VCs are deploying capital into new L2-native protocols. But look at the drillers: the number of L2 projects actively seeking mainnet launch has more than doubled in the last quarter. Every new L2 is a “pipeline” that needs to be filled with transactions. If all of them succeed, we are not heading toward a world of infinite scaling. We are heading toward a world where the aggregate demand for block space once again outstrips supply, and fees — even on L2s — start to rise. The market is pricing the relief, not the return of congestion.
Based on my experience auditing DeFi protocols during the last cycle, I can tell you that the most dangerous moment is right after the bottleneck is removed. That is when everyone rushes in, assuming the problem is permanently solved. They forget that every relief valve becomes a faucet. The Permian drillers forgot. The Ethereum ecosystem is at risk of forgetting too.
So what do we do? We do not stop building. We double down on horizontal scaling — migrating to even more efficient storage schemes, incentivizing L2 diversity, and preparing for a world where each L2 becomes its own ecosystem with its own demand curve. But we must stop telling ourselves that the scaling problem is behind us. It is not. It is simply taking a different shape. The next phase of this bull market will not be defined by how low fees can go. It will be defined by how gracefully we handle the moment when demand catches up to supply again.
When the next gas spike hits — and it will — do not be surprised. Remember the pipelines. Remember the drillers. And ask yourself: what are we building right now that will become the next bottleneck?
