A few weeks ago, a tweet from a Chinese AI researcher named Zhu Huajiang sent shockwaves through Silicon Valley's machine learning circles. He argued that in the era of frontier models, the infrastructure engineer has been treated as a second-class citizen — a 'peasant' beneath the 'noble' research scientist. Elon Musk resonated with that sentiment, calling the hierarchy 'toxic.' This wasn't just a tech squabble. It was a narrative shift with profound implications for how we build complex systems. And if you think this has nothing to do with crypto, you're missing the signal.
For years, crypto's own hierarchy has been equally rigid. The 'researchers' — the ones publishing papers on novel consensus mechanisms, zero-knowledge proofs, or sharding architectures — sit at the top. Beneath them are the 'engineers' who implement, optimize, and ship. We worship whitepapers. We fund PhDs. We treat infrastructure teams as a cost center, not the core of innovation. But the AI culture war reveals a blind spot: in both fields, the velocity of iteration now depends on the people who write the low-level code, not the ones who dream up the grand theory.
Reading between the code to find the human story. I've been tracking this pattern since my days as a Token Fund Investment Manager in Zurich, during the DeFi Summer of 2020. Back then, I noticed that teams with flat organizational structures — where engineers could directly refactor smart contracts without waiting for research approval — consistently launched faster and captured more liquidity. Uniswap's tiny team shipped V2 and V3 with ruthless efficiency. Meanwhile, projects like Synthetix, with a heavy layer of academic tokenomics, often got stuck in debate cycles. The narrative was clear: engineering culture, not research pedigree, was the real moat.
Fast-forward to today. The AI debate mirrors exactly what I'm seeing in Bitcoin Layer 2s. Over 80% of projects claiming to be 'Bitcoin L2s' are actually Ethereum-style rollups or sidechains that have rebranded to chase hype. The real Bitcoin community — the Core developers, the OG cypherpunks — doesn't even acknowledge them. Why? Because Bitcoin's culture is deeply conservative, almost anti-research. It values minimalism, security, and engineering simplicity over theoretical novelty. Unearthing value where others see only chaos means recognizing that culture is the ultimate determinant of which protocols survive the bear market.
Let me ground this in my own framework: 'Narrative Velocity Tracking.' In 2022, during the Luna collapse, I observed a similar dynamic. Terra's 'researcher-first' culture produced the algorithmic stablecoin theory, but its engineering infrastructure was brittle. The very people who could have stress-tested the system were undervalued. Contrast that with a project like Chainlink, which has always been engineering-led, iterating on oracles through incremental improvements rather than moonshot papers. Chainlink didn't have a grand narrative; it had reliable code. And it's still standing.
Now, the contrarian angle. The AI pundits who defend hierarchy argue that deep theoretical work requires isolation and status. They say you can't do foundational research without a 'noble class' free from the grunt work of debugging CUDA kernels. There's a kernel of truth there. In crypto, we've seen similar — the deepest innovations in zero-knowledge proofs came from academic labs like the ones behind zkSync and Starkware, which maintained a clear separation between researchers and engineers. But look at the timeline: those projects took years to ship a mainnet. Meanwhile, an engineering-heavy team like Solana shipped a blazing-fast L1 in months, albeit with reliability issues. The trade-off is real.
My take? The pendulum will swing again. As the base infrastructure matures, the next wave of breakthrough research — say, fully homomorphic encryption or post-quantum signatures — may once again require deep specialization that flat teams struggle to provide. But for now, in this sideways market, the advantage goes to the builders, not the theorists. Chop is for positioning. The teams that are quietly optimizing their execution layers, slashing latency, and reducing gas costs are the ones accumulating liquidity. The teams still arguing about tokenomics models are bleeding LPs.
What does this mean for your portfolio? Look for projects where the GitHub commit history is longer than the whitepaper. Favor teams that talk about 'training frameworks' over 'paradigm shifts.' Ask yourself: is this team organized to iterate fast, or to impress academics? The next narrative in crypto isn't about a new chain or protocol — it's about how the team is wired. The ones with flat, engineering-first cultures will pull ahead. Reading between the code to find the human story is how you spot them before the market catches on.
One final thought from the AI debate: Musk's response was more than a critique of Silicon Valley's ego. It was a reminder that in any complex system, the people who make the infrastructure hum are the ones who deserve the highest status. In crypto, we've been too busy worshipping VCs and research papers. The engineers have been quietly building the rails. Now they're coming for the narrative throne.
