A look at some US energy companies linked to AI infrastructure, such as Sam Altman-backed Oklo, whose stocks have soared despite having little to no revenue
Concept stocks with no revenue have soaring valuations — Forget about the froth in tech valuations. The real excess might be building up in energy stocks. X: @aaronback . LinkedIn: Gerard Reid and WSJ Heard on the Street Bluesky: @tprstly.com , @kostyack , and @pauljdavies X: @aaronback : A group of “pre-revenue” energy stocks has ballooned to a $45 billion market value in hopes that AI companies will someday buy their power. It's everything from nano-reactor concepts to refurbished, used gas turbines. https://www.wsj.com/... LinkedIn: Gerard Reid : There is no doubt that #ai will revolutionise our world and there is also no doubt that AI needs lots of electricity but whether … WSJ Heard on the Street : A group of “pre-revenue” energy stocks has ballooned to a $45 billion market value in hopes that AI companies will someday buy their power. … Bluesky: Theo Priestley / @tprstly.com : Lots of people saying that quantum is the next big grift but it's not - it's energy. Specifically, fusion startups promising unlimited clean energy for AI data centres. [embedded post] John Kostyack / @kostyack : To understand the AI and data center boom, it's important to know both the role of “zero-revenue” companies and the opaque and largely unregulated Wall Street financing vehicles that make them possible. [embedded post] Paul J Davies / @pauljdavies : Good piece on multi-billion dollar *zero-revenue* speculative energy companies, incl Sam's Oklo and one run by the ex-CEO of a failed anti-woke bank >> — Heard on the Street: Forget about the froth in tech valuations. The real excess might be building up in energy stocks. www.wsj.com/tech/ai/ai-b...
Context & Ripple Effects
AI demand had already broadened the market’s focus from chipmakers to utilities, energy and materials needed to support AI systems. This coverage narrows that shift to earlier-stage power suppliers whose valuations rest on prospective data-center demand rather than operating revenue.
It also sits beside reporting on experimental clean-energy projects pursued for AI power needs, showing how the infrastructure buildout is pulling both established and unproven generation options into the same investment narrative.
First-order effects
- Oklo and other pre-revenue power concepts gain market valuations and investor attention based on their perceived fit with future AI infrastructure, despite limited evidence of current commercial revenue.
- The $45 billion valuation assigned to this group makes future AI power purchasing a central benchmark for how investors assess nano-reactors, fusion ventures and refurbished gas-turbine propositions.
Second-order effects
- Established utilities and energy suppliers face a more crowded investor market for the AI-power theme, as capital is also directed toward companies promising more specialized or novel generation capacity.
- AI developers and data-center builders gain a wider set of prospective power partners, but the contrast between elevated valuations and limited revenue raises execution risk around which projects can actually supply power.
Third-order effects
- If AI power demand remains the organizing investment thesis, electricity generation could be financed increasingly as a strategic extension of compute infrastructure rather than as a standalone utility market.
- The pattern could widen the gap between companies that secure credible power-delivery commitments and those valued mainly on projected AI demand, making project execution a stronger determinant of industry structure.
The trend: AI infrastructure financing is expanding beyond chips and data centers into a speculative contest to own the power capacity that could serve them.