Ethereum Marketed as AI Infrastructure Play

Tom Lee’s view is built on that same premise. The Fundstrat co-founder has argued that Ethereum can emerge as a winner in the AI trade, a call that comes as semiconductor names have lost momentum and the SOXX ETF has fallen sharply from its spring highs. The fund was trading near $523 on July 17, well below its 50-day moving average of about $566, a sign that the chip trade has cooled even before any true broad-based collapse in earnings. Nvidia, the bellwether for AI spending, has also slipped from recent levels and is trading just above its 200-day moving average, suggesting the market is becoming more selective about which AI beneficiaries deserve premium valuations.

Ethereum, by contrast, is showing the kind of violent volatility investors expect from a maturing but still deeply speculative asset. It recently traded around $1,843, far below its 200-day average of roughly $2,195, even after a strong rebound from an early-year washout. Its 50-day moving average sits near $1,741, and the token is hovering just above that level, which tells you the market is still deciding whether this is a durable recovery or another tradable bounce. Standard technical measures such as RSI readings and the MACD have improved from oversold conditions, but they do not change the bigger story: Ethereum remains a high-beta asset whose fortunes can swing quickly with sentiment around AI, liquidity and macro risk.

Why should long-term investors care? Because the AI winner may not be the obvious one. The first wave of AI investing rewarded chipmakers, cloud providers and the hardware stack. But if the build-out keeps expanding, capital often migrates toward the rails, standards and settlement layers that sit underneath the hype. That is where Ethereum bulls are planting their flag. They see a network that already has scale, developer mindshare and a growing role in tokenization and digital asset infrastructure. If AI agents, machine-to-machine payments and on-chain settlement become more important over time, the winners may be the networks that can host them, not just the companies selling shovels.

There is a catch, of course. Ethereum has spent years frustrating investors who bought the story but could not stomach the drawdowns. Citi’s recent cut to its outlook after several weak quarters is a reminder that narrative alone is not enough. The token still has to prove it can translate adoption into sustained economic value, and it has to do so in a market where capital rotates fast and macro shocks can overwhelm fundamentals in the short run. Escalating geopolitical tensions and a stronger dollar can also sap appetite for risky assets, especially those without cash flow.

Still, the bigger investment lesson is hard to miss: the AI trade is broadening. The market is no longer asking only which chipmaker can sell the most accelerators. It is asking where the durable infrastructure will sit if AI becomes as foundational as the internet itself. That is why Lee’s Ethereum call is interesting even for investors who never plan to own a single token. It signals that the next phase of the AI boom may be about networks, settlement, and ownership — not just semiconductors.

For investors with a long time horizon, Ethereum is still best treated as a small, high-risk satellite position rather than a portfolio centerpiece. But if you believe the AI build-out will last for years, and if you are looking for assets that could benefit from the convergence of crypto infrastructure and machine-driven commerce, Ethereum belongs on the watchlist. In markets like this, patience and diversification still win.