Helion
Helion
One-line summary: Sam Altman-backed private fusion company whose 50 MW Orion plant cleared Washington State regulatory approvals in June 2026 — positioned as the first commercial fusion plant, contracted to supply Microsoft starting 2028. Not tradeable directly; matters as a dated milestone in the AI-power-gap timeline.
What it is
A pulsed-fusion startup (deuterium + helium-3, field-reversed-configuration plasmas) whose architecture directly recovers electricity from the plasma's changing magnetic flux via Faraday induction — skipping the heat → steam → turbine chain of tokamak designs. From 2026-07-01-podcast-moonshots-sonnet-5-drops-fable-5-will-return-fusion-s-first: raised ~$1B at a $5.4B valuation; Sam Altman was the largest early investor and stepped down as executive chairman ~two months before this episode (per Diamandis, freeing Helion for OpenAI-scale partnerships).
Key facts (all from 2026-07-01-podcast-moonshots-sonnet-5-drops-fable-5-will-return-fusion-s-first)
- peter-diamandis: "On June 16th... Helion cleared the required Washington state regulatory approvals for its Orion fusion power plant which is intended to supply Microsoft with 50 megawatts of power starting in 2028."
- Scale caveat, same speaker: 50 MW is small — "we talk about gigawatt level plants, this is 50 megawatts." Commonwealth Fusion (the other leader; also ~$1B raised) targets its first 400 MW plant ~2032.
- alexander-wissner-gross: the fusion "triple product" (density × confinement time × temperature) has improved steadily for 50 years — "you're watching the triple product, you could predict when this is going to happen and it's imminent."
- Sector context: ~50 privately funded fusion companies, ~$6B raised in aggregate.
Why it matters to stock-market
- Timing, not displacement. Even the techno-optimist source concedes fusion (and SMR/Gen-3 fission) won't arrive at scale "till the early to mid-2030s" (peter-diamandis). So the Helion milestone does NOT dent the 2026–2030 window of nuclear-baseload-for-ai-data-centers / pjm-capacity-prices-to-nuclear-premium — existing fission fleets and the grid carry the AI load through the decade.
- Watch as falsifier fuel. If Helion delivers 50 MW to Microsoft on time in 2028 and mass-manufactures, the long-dated terminal value of uranium/fission names (CCJ, CEG, UEC) starts to compress; a slip reinforces the fission premium.
- Skepticism note: Moonshots is promotional and Diamandis/Blundin are ecosystem investors; the regulatory clearance is a state-level approval, not an NRC license, and "first fusion plant" remains a 2028 promise from a pre-revenue private company.