Five Signals From the Frontier
Packy McCormick's 182nd Weekly Dose of Optimism rounds up five stories spanning energy storage, fusion power, psychedelic medicine, fintech infrastructure, and rare disease research. Taken individually, each is noteworthy. Taken together, they sketch the outline of what McCormick clearly believes is an acceleration in technological progress across nearly every domain that matters.
Iron-Air Batteries and the BYOE Era
The headline item is a billion-dollar order. Google has contracted Form Energy, a nine-year-old startup, to supply a 30 gigawatt-hour battery system for a planned data center in Pine Island, Minnesota. The deal is part of what McCormick frames as the hyperscaler push to "Bring Your Own Electricity."
Iron-air batteries are low power density and slow response, but extraordinarily cheap on a per-kWh basis, roughly 10% of the cost of lithium-ion. They're perfect for longer-term storage, up to 100 hours, for multi-day lulls in sun and wind.
The chemistry itself is elegantly simple. McCormick walks through how iron-air batteries work through a process akin to reverse rusting: iron oxidizes to discharge electricity, then gets electrochemically reduced back to metallic iron when charged. The raw materials are cheap and abundant. Form will manufacture the batteries at a former steel mill site in Weirton, West Virginia.
This is good news because we love batteries here at Not Boring, and because it's a rare win for a western battery company in a category that's been dominated by China.
That framing deserves a gentle counterpoint. A single billion-dollar order, however impressive, does not constitute a shift in global battery manufacturing dominance. China's CATL and BYD produce at scales that dwarf anything Form Energy has demonstrated. The real test will be whether Form can manufacture reliably at volume, on time, and at the projected cost. That story has not been written yet.
Stellarators Get Serious in Bavaria
Proxima Fusion, a German startup spun out of the Max Planck Institute, signed a memorandum of understanding to build what would be the world's first commercial stellarator fusion power plant. The project needs roughly two billion euros for a demonstration facility, with Bavaria and the German federal government potentially providing up to 1.6 billion euros combined.
We made Tokamaks first not because they were best, but because they were the easiest to design and manufacture. Stellarators, with their weird twists, were harder to design and manufacture, but closer to the platonic ideal of a fusion generator.
McCormick is characteristically enthusiastic, calling it a "very cool way to get German energy back on track" after the country's decision to shutter its nuclear fleet. But he also acknowledges the distance between an MoU and electrons on the grid.
Of course, it's just an MoU. The plant still needs to be built, and then there's the tricky matter of achieving Q>1 and, eventually, Q> whatever it needs to get to to be economically viable.
That qualifier is doing heavy lifting. Fusion timelines have a long and humbling history of slipping by decades, and stellarators, despite their theoretical elegance, remain unproven as commercial power sources. Government MoUs have a way of evaporating when budget cycles turn.
Psilocybin Pushes Past Depression
Two studies dominate the psychedelic medicine section. Compass Pathways achieved its primary endpoint in a second Phase 3 trial for COMP360 psilocybin in treatment-resistant depression, making the company three for three on trials. A rolling FDA approval submission is planned for late 2026.
The more intriguing result, however, came from Johns Hopkins.
A Johns Hopkins pilot study of 20 adults with well-documented post-treatment Lyme disease found that psilocybin, given with psychological support, produced significant and lasting reductions in multi-system symptom burden, including improved mood, fatigue, sleep, pain, and quality of life, that persisted for up to six months.
McCormick draws a sharp inference from this: if psilocybin helps with a chronic neuroimmunological condition, the mechanisms at work likely go beyond emotional processing in therapy. He also makes a candid observation about the pharmaceutical business model.
From a pharma business perspective, one of the biggest challenges with psilocybin is that it just works. There's no opportunity to keep patients on (and paying for) meds for the rest of their lives like there is with SSRIs.
That is a striking claim, but it bears noting that the Lyme study had only 20 participants and no control group. Pilot studies of that size generate hypotheses; they do not confirm them. The durability argument is suggestive, not conclusive.
Stripe's Letter as Economic Observatory
Patrick and John Collison published Stripe's 2025 annual letter, and McCormick treats it as one of the best pieces of economic writing available, largely because Stripe processes roughly 1.6 percent of global GDP.
Their 2025 cohort of new businesses is growing 50% faster than the 2024 cohort. The number of companies hitting $10 million ARR within three months of launch doubled. iOS app releases jumped 60% year-over-year in December. GitHub pushes surged 41% between Q3 2024 and Q3 2025.
Three sections of the letter get particular attention. The first is a five-level framework for "agentic commerce," ranging from agents that fill out checkout forms to agents that anticipate needs and buy autonomously. The second covers stablecoins: payment volume doubled to around 400 billion dollars, and Stripe is building a payments blockchain called Tempo with Paradigm. The third is a closing argument about what the Collisons call a "Republic of Permissions."
Technologies succeed or fail not just on their merits but on whether the web of regulators, committees, and courts lets them through.
McCormick clearly finds the regulatory argument the most resonant. He quotes the letter's closing image approvingly.
We're reminded of the phenomenon of falling into a large black hole... We write this letter at what may well turn out to be the advent of a different and hopefully much more beneficent singularity.
Vitamins as Genomic Medicine
The final story comes from the Arc Institute, published in Cell. Researchers ran a genome-wide CRISPR screen to identify genetic diseases that might respond to simple vitamin supplementation. The top hit for vitamin B3 was NAXD disease, a condition caused by mutations in a repair enzyme essential for redox biology. It is typically lethal in early childhood.
In knockout mice, adding vitamin B3 to food from birth increased lifespan more than 40-fold. McCormick, channeling Arc researcher Ulkar Aghayeva, frames the finding as proof that vitamin biology still contains unexplored territory despite all 13 classical vitamins being discovered by 1948.
The question the piece leaves hanging is a good one: how many other rare genetic diseases might yield to interventions this simple? The answer, almost certainly, is more than zero and fewer than we would hope.
Bottom Line
McCormick's optimism roundups work because they aggregate real signals rather than speculative hype. A billion-dollar battery order is a fact. Three successful Phase 3 trials are facts. A 40-fold lifespan extension in mice is a fact. The editorial choices lean bullish, and some of the extrapolations outrun the evidence, but the underlying data points are genuine. For readers tracking the pace of technological change across energy, medicine, and infrastructure, this installment offers a useful and largely honest snapshot of where things stand in early 2026.