Is Crypto Broken? | Roundup

Miles argues that the current crypto bear market sentiment is overstated relative to 2018 and 2022, and that this is actually the least confusing moment the industry has ever been through, with early existential questions resolved and value creation now the central remaining challenge. Crypto appears to be the only technology sector currently in a bear market while AI is entering a bull market, and the concern raised is that AI may eventually need to work through the same excess-capital hangover that crypto is experiencing now. The 2017 ICO era is cited as the reference point, with EOS and Block One raising 4.5 billion dollars as one example of enormous capital misallocation underwritten on false priors and speculative infrastructure projections rather than fundamental analysis. Pomp tweeting that 20 percent of the world's billionaires were crypto people is identified as a major warning signal in hindsight.

The incentive structure of the industry was oriented toward token launches around speculative narratives rather than business fundamentals, with VC calls dominated by questions about token timelines rather than value creation. This discouraged real business builders and produced very few household names outside of Bitcoin and Ethereum. Founders in the current cycle are following capital into already competitive spaces like prediction markets and stablecoins rather than pursuing non-obvious uncontested ideas, and VCs are pushing startups toward established fintech niches where they are poorly positioned against large incumbents. The bar for investing in crypto outside of AI is described as higher than it has ever been.

Value capture has risen up the application layer, meaning app-layer builders from two or three years ago are better positioned now than infrastructure-focused teams. Stablecoins are seeing the most adoption, particularly in emerging economies, but their rise primarily benefits chains with distribution like Tempo and Circle rather than legacy layer-one networks. A piece by David Hoffman argued that layer-one assets are increasingly valued on revenues and fees rather than functioning as money, and the thesis that assets like ETH would function as money is now being proven wrong. Ethereum is described as being in a particularly difficult position partly because stablecoin adoption does not favor legacy layer ones.

Zave argues that AI is harming crypto more than benefiting it in the short term and specifically expects AI to hurt DeFi security over the next six to twelve months before conditions improve. He remains long-term bullish on AI agents interacting with blockchain as a potential next cycle catalyst but expects significant near-term risk. Every prior crypto cycle required a new catalyst such as DeFi or layer ones to exit a trough, and 2026 is expected to be a challenging year as the industry waits to identify what that catalyst will be. Crypto is expected to be among the last beneficiaries of the AI wave once multi-agent infrastructure matures.

The panel identifies a structural split forming in real time between an institutional camp and a cypherpunk camp oriented around the industry's original principles, with one speaker reframing this as scaling versus new ideas. Institutions are described as building more crypto companies than cypherpunk-oriented founders for the first time, though the speakers acknowledge having no hard data to support this observation. Zcash rallying after more than ten years of existence is cited as concrete evidence of a cyberpunk revival, and privacy-related applications are described as a potentially viable market. Quantum computing is expected to arrive within the current crypto cycle and is flagged as relevant to cryptography concerns. One speaker notes that a return to cyberpunk ideals would likely represent a smaller market than the institutional path.

The regulatory environment is described as having fundamentally changed, with the president and secretary of treasury stating they want to move dollars on-chain and both the SEC and CFTC stating they want to rebuild finance on-chain. One speaker describes the past six years of DAO governance combined with US regulatory hostility as building with two hands tied behind your back, making it remarkable that anything meaningful was built at all. One speaker predicts that businesses building in crypto now cannot fail to make significant money given these political and regulatory catalysts, though the GENIUS Act has not yet been implemented and is not expected until November of this year.

Pre-seed crypto funding is described as nearly non-existent for category-defining ideas without revenue. The talent pool is characterized as a historically small Venn diagram of domain experts and crypto industry experts, with ideologues, cypherpunks, grifters, and academics largely pushed out. The panel frames novel long-duration primitives comparable to Uniswap or MakerDAO from 2017 as what founders should be aiming to build now, and expects new categories to emerge in the current cycle beyond known ones like prediction markets, stablecoins, and exchanges.