Hook
In May 2024, Masayoshi Son placed a $155 billion bet on Helion Energy, a company that promises to deliver nuclear fusion—the holy grail of clean energy—within 15 years. The narrative is seductive: AI’s insatiable appetite for electricity will be solved not by incremental solar panels or batteries, but by a technological singularity that rewrites the laws of physics. But as a crypto educator who has spent years auditing whitepapers and chasing promises, I see a familiar pattern: a charismatic founder (Son), a complex black-box technology (Helion’s magnetized target fusion), and a timeline that conveniently aligns with the end of the current investment horizon. The crypto world knows this script all too well.
Context
SoftBank’s Vision Fund has a history of betting on narratives that sound too perfect to be true—WeWork was supposed to disrupt real estate, Uber was going to eliminate car ownership. Now, Son is pivoting to energy, arguing that AI data centers will demand 3 terawatts of power by 2040, and only fusion can deliver that without destroying the planet. Helion uses a deuterium-helium-3 fuel cycle, which promises near-zero radioactive waste. But 3He is almost impossible to source on Earth—it comes from nuclear warheads or, hypothetically, the Moon. The parallels to crypto’s early days are unmistakable: a revolutionary claim, a complex mechanism that few understand, and a call to “trust the process.” In decentralized finance, we learned that trust is not a strategy—verification is.
Core: The Fusion Whitepaper That Won’t Pass Our Audit
Let’s apply the same scrutiny we use on DeFi protocols to Helion. First, the technology readiness level: Helion has not demonstrated a net-energy-positive prototype at any scale. In crypto, this would be like launching a token without a working mainnet—a pure speculation vehicle. The claim that fusion will be commercial by 2040 is based on a linear extrapolation of R&D progress, ignoring the “valley of death” that kills 99% of hardware startups. Based on my experience auditing ICO whitepapers in 2017, I learned that when a project’s timeline relies on “breakthroughs” rather than incremental milestones, the probability of delivery is near zero.
Second, the fuel supply. Helion’s reliance on 3He is akin to a blockchain that depends on a single, unregulated oracle. Today, the global supply of 3He is less than 200 kilograms per year, mostly from U.S. nuclear weapons stockpiles. Scaling to commercial power would require either a massive new production chain (neutron bombardment of lithium) or extraterrestrial mining. Both are technological wildcards. In crypto, we penalize projects that outsource critical functions to unverifiable off-chain sources. Helion does exactly that.
Third, the cost narrative. Son implies fusion will be cheaper than renewables. But even if Helion succeeds, the levelized cost of electricity (LCOE) from fusion is a complete unknown. The International Energy Agency’s most optimistic projections for fusion are 50-100% higher than current solar-plus-storage costs. Meanwhile, solar LCOE continues to drop by 5-10% annually. By 2040, solar will likely be $0.02/kWh or less—a moving target that fusion cannot outrun. This is the same trap that beset many ICOs: assuming a fixed future cost for a new technology while ignoring exponential improvements in existing ones.
Contrarian: The Distraction Is the Real Product
The contrarian angle is not that fusion will fail—it might succeed eventually—but that Son’s narrative actively harms the energy transition. By positioning fusion as the only true solution for AI’s energy needs, he implicitly devalues immediate, deployable solutions: natural gas, renewables, long-duration storage, demand response, and efficiency. This is reminiscent of the “ETH killer” narrative in crypto—every new chain claimed to be the final solution, while Ethereum itself continued to improve. The result was a misallocation of capital and attention.
Furthermore, the article’s framing—AI needs fusion—creates a false binary: either we accept natural gas (dirty) or wait for fusion (clean). The middle ground of renewables-plus-storage is ignored. But the data is clear: every major AI company is already signing long-term power purchase agreements (PPAs) for solar and wind. Google, Microsoft, and Amazon have committed to 100% renewable energy by 2030. They are not waiting for fusion. The real risk is that Son’s “moon shot” narrative slows down necessary investment in grid modernization and storage infrastructure.
Takeaway
SoftBank’s fusion bet is not an investment—it is an act of belief. And belief, as we know in crypto, is not a substitute for verification. The ledger of reality will not be kind to Helion if it fails to deliver net energy gain by 2030. Until then, Son’s words are white papers without code. As I tell my students at BlockMind Academy: “Truth is not consensus, it is verification.” The same applies to energy. “The ledger remembers what the crowd forgets”—and the crowd has already forgotten that every previous fusion “breakthrough” in the last 50 years was followed by a decades-long slog. “We build walls of code to protect hearts of flesh,” but Son is building walls of hype to protect a balance sheet. The future belongs to those who audit the present—and the present says: deploy renewables, build storage, and treat fusion as a long-term public good, not a private lottery ticket.