The Sienna AI Simulation Lineage

The first of the Ten Technologies was Technology 5: The Virtual Social Business Network, where in 2003 we created the world’s first commercial virtual city tour. In 2012, we built upon this with Technology 6: The gaming simulation UCS, including Š🌀ŔÉŚ, its name inspired by Asimov’s Psychohistory: the Universal Colonisation Simulator.

2012 was a year of hyper-creativity: just one month later came the Predictive Quantum Software (PQS), which borrowed from the Manhattan Project’s Monte Carlo effect as an economic simulator — four years before DeepMind borrowed similar logic to create AlphaGo. For the next four years we developed Technologies 1–4, while continuing to study quantum theory.

In 2016, we created M-Systems: an economic theory of everything.

M-Systems added to the UCS the Feynman “sum over histories” formulation of quantum mechanics, together with pruning techniques such as renormalisation. This was where the inspiration for BrainSIM first materialised — though originally applied to macroeconomics, not neurology.

Years of simulation work on how to change the fortunes of the Global South, and particularly the Malawi History 3 simulation, led us to argue an economic technology that could boost Malawi from near-zero to 1% of global GDP per capita — the equivalent of Norway or the UAE. From this lineage came the 87 Quintillion Histories, and in 2021 the design was simplified into the Ten Technologies (T10T).

Out of Technology 6 came three macro-technologies:

• 7 — Š🌀ŔÉŚ growth multiplier
• 8 — NetZero Dynamic Comparative Advantage
• 9 — Grand Spin Networks

The simulations that form the bedrock of these macro-technologies directly inspired the BrainSIM Project, and when we put this design to GPT-5, Gemini, and Grok-4, all three surprisingly agreed: this is something that could really happen.

And so the story loops back: when you’ve worked with systems designed to simulate 87 quintillion possible economic futures, reducing the search space to “just a billion” in neurology is almost modest. But it’s enough. A billion is enough.

A Practical Case for BrainSIM

In 2100i12E5) The BrainSIM Project — Specifications, Proofs, Costing & Resourcing Plan (link), we set out the detailed case. Here is the essence in plain language:

• Baseline today (2025): Across major neuro conditions (AD, PD, CSS, MS), even elite programs only manage 10–100 protocol variants per year to a decision point.
• What is a “variant”? A fork of a treatment recipe (drug(s) + dose + timing + combinations ± stimulation/rehab) tuned for a disease subtype.
• What is an “actionable decision-test”? A go/modify/stop decision in one context (e.g., dose, subtype, safety gate). A single variant spawns many decision-tests. By 2029 we aim for 10^5–10^6× more decision-tests.
• AlphaFold analogy: AlphaFold did 100–1,000× faster protein structures. BrainSIM applies the same logic to brain protocols.
• Throughput ≠ cures: Billions× exploration doesn’t mean billions× cures. It means fewer dead ends and faster discovery. Biology and regulation still set the speed limit.
• How it works: Generate thousands of variants → prune fast → elevate winners → confirm safety in short, single-patient trials. Built-in “smoke alarms” predict side-effects like nausea.
• Numbers:
  • Baseline: 10–100 evaluations/year today
  • 2029: 10^5–10^6× → millions/year
  • 2031: 10^7–10^8× (and 10^9× plausible with neuromorphic/exascale)
• Timelines (estimates, not promises):
  • By 2035: ~70–85% odds of management-level relief in CSS/PEM, ~60–75% in PD/AD
  • By 2040: ~80–95% management, ~50–70% CSS cure, ~30–50% PD/AD/MS
  • By 2045: ~90–99% management universal, ~50–70%+ CSS cure, ~30–50%+ PD/AD/MS
• Costs: <$50M for a flagship program. Compute <2% of a GPT-7 training run. ROI dwarfs pharma R&D norms.

The Economic Bridge

BrainSIM is not another speculative research program. It is the direct medical application of simulation engines already proven across economics. By 2029 we expect millions× throughput; by 2031 plausibly billions×. Half of those simulations will optimise what we already know; the other half will be “long shots” — the place where penicillin-class surprises live.

This is why neurology is uniquely positioned to prove AGI’s worth to humanity. And it is why a small slice of frontier AI infrastructure — just 1% — could deliver the most important public health breakthrough since antibiotics.

For readers who want to step beyond medicine into the economic foundation, see: The Economics of AGI.

Author’s Note

The other piece of luck for neurology — and decisive misfortune for me — is that I have been struck down with iatrogenic Small Fibre Neuropathy (SFN), Central Sensitisation Syndrome (CSS), Post-Exertional Malaise (PEM), and Fibromyalgia (FM phenotype). A neurological firestorm of pain.

If you look at my work since 2011, no one would deny the altruistic intent. But this goal is also purely selfish. My survival depends on creating this technology. I, Nick Ray Ball, founder of Sienna AI and the inventor of the Economics of AGI, am not going to stop. Whether this is championed by me or by others, this will happen. One way or another, I will either do it — or die trying.

You don’t get more devotion to a cause than that.