Aaron Kushner

Aaron Kushner, Ph.D.

Making AI Deliver.

AI doesn’t solve all our problems. Just half. Giving us the space to solve the other half.

City of Oakland Bentley School The College Preparatory School Cal — Oski the Bear Theravance Biopharma, Dublin Laney College UC Irvine — Peter the Anteater The Hitchhiker's Guide to the Galaxy — Don't Panic

Hi! My AI-powered research is focused on driving maximum AI diffusion into the general non-technical public at maximum safe speed. This involves providing not just technical empowerment tools and products for the individual non-technical user, but also across the position profile from large institutions to small businesses. Additional work includes education on how to understand AI, how it works, and how to use the tools to maximize its benefit, as well as research toward affordable solutions to long-standing socio-technoeconomic problems that are now affordably and equitably addressable — as well as the problems AI created ;). And, of course, building my own dreams into viable businesses!

The CS-9 aircraft in side view, wearing a Fable 5 decal, riding a green ribbon of CFD streamlines

If you are interested in collaborating on scientific communications and/or AI product development, or would like to engage my AI implementation expertise on a consulting basis, please reach out through the contact form below.

Résumé (PDF) ↓ Read it in the browser →


Products/Projects

Flagship Product

General Research & Development Workbench

Built-in AGI-adjacent behavior: memory, judgment, verification, continuity.

Capability is not the constraint — reach is Three cards joined left to right by arrows. Today: the machine runs on an IDE, pictured beside the label; a funnel narrows to a gate marked "gate: the IDE" and almost none of the hundred-cell population grid is filled; it cannot be distributed without losing the intellectual property. In testing: the web-based R&D workbench, running on a web-based Linux VM and pictured as its live browser cockpit; its gate is marked "designed for R&D" and about eight cells are filled, the share of adults holding a STEM bachelor's degree or higher; it is infinitely distributable with the IP intact. A dashed arrow marked "port in progress" leads to the third card: after the port the same machine runs on a simple interface, pictured as an early personal computer showing the word hello, there is no gate, and the entire grid is filled. {"id":"b73f20","program":"akushnerphd-me","lane":"akushnerphd-me","subject":"workbench","type":"figure","created":"2026-07-24","title":"The constraint is reach, not capability - three states (Opus 5 blind, edited)","provenance":{"session":"683eba5b","derived_from":"workbench-opus5"},"tags":["workbench","flagship","product","reach","openrouter","distribution"],"schema":1} PORT IN PROGRESS TODAY My personal human-like collaboration AI R&D machine RUNS ON IDE GATE: THE IDE WHO CAN REACH IT A SLIVER Reachable only by people comfortable with a developers environment. Not distributable without IP loss. IN TESTING The web-based R&D workbench RUNS ON Web-based Linux VM GATE: DESIGNED FOR R&D WHO CAN REACH IT ~8% Reachable by adults holding a STEM bachelor’s or higher. Infinitely distributable · IP protected. AFTER THE PORT — SAME MACHINE Universal “Doing Machine” for anyone with internet and OpenRouter RUNS ON Simple, Intuitive UI, personalized AI Colleague/Companion NO GATE — ANY BROWSER WHO CAN REACH IT EVERYONE Reachable by all people.

A general R&D environment built on Claude Code: a behavioral layer of memory, judgment, verification, and continuity that turns a frontier model into a persistent collaborator. Proven in daily use on one user — every project on this page is its output — with a hosted, person-agnostic version built and in testing.

What it is

  • The harness around the model: persistent memory across sessions, continuity that survives process death, and verification gates that block “done” without observed evidence.
  • Adapted from Claude Code’s customizable core — a coding tool tuned over months of daily use into a general research workbench.
  • The target has a market-legible name: Personal AGI — models and behavior uniquely tailored to each individual user, person-agnostic by design from the project’s first weeks.

Proven, and what’s next

  • Built to adapt to one person — it worked: context, standards, and working style carried across sessions — the built-in hashed provenance record is the measurable proof of this claim.
  • Hosted workbench for institutions and independent researchers: built, in testing. Roadmap: the “iMac moment” layer that makes it a product for non-technical users.

Tokenomics — the institutional application

  • A deployable answer to the AI-spend measurement gap the business press now calls tokenomics: companies buying AI with no accounting practices to measure the return.
  • Every seat’s output flows upward as a synthesized per-seat report — type, scale, and impact of granular token spend, traceable to the commit; the synthesis repeats seat → team → corporate.
  • At least one human signs every final call: the push-button generator exists, and its first real report (a whole-portfolio week, 400 commits) renders draft — unsigned until a human signs it.
  • Removes manager–builder friction in both directions: seats stop burning loaded time on manager reports; managers stop dropping valuable work to synthesize raw ones.

Air Sports & the Engineering Engine

The AirPark Digital Twin

Pre-AI: $250,000, a team of engineers, one year. Post-Fable: $5,000 plus Claude Max, one non-engineer with our workbench, three months.

An air-sports venue and its engineering program, built as one digital twin. The AirPark is guaranteed-safe flight for everyone — licensed or not — and the aircraft program behind it is the seed of a universal, natural-language engineering engine.

The venue

  • Fly in and step into a high-performance machine: aerobatics, engine-outs, stall/spin recovery — all backed by an autonomous fail-safe. No navigation, no pre-flight.
  • No license required inside the protected zone: guided first flights, AR race gates at altitude, coached mock dogfights.
  • The full-park digital twin flies millions of simulated hours against worst-case scenarios before any part of the park is built.
  • Purpose-built electric aircraft: austere, rugged, a fraction of GA cost — where private pilots can train landings, engine outs, and deep stalls with high-tech autonomous escape backup.

The engineering engine

  • Every airframe hardens a universal engineering loop: natural-language design intent → CFD, structural FEM, flight simulation, crash test — physics as the pass/fail oracle, “deploy” as a manufacturable package.
  • The systems side — digital twin, AI trainer, cybersecurity, collision avoidance — is integrated from inception with the physical side: complete cyberphysical systems.
  • Real deployment risks remain (engineers, IP, real estate, the FAA) — but the sunk cost-to-prototype-to-Series-A engineering risk is dramatically reduced.
  • Generalizes to an open, hard-tech-agnostic atoms-to-objects engine operable in natural language — de-risking hard-tech ideas that previously never found funding.

Independent Policy Research

Responsible AI Diffusion

Two substrates, different in kind — permanently paired Left, a living brain with a zoom callout of networked nerve cells: the chemical substrate, lived through time, staking a finite life. Right, a microchip with a zoom callout of a mathematically weighted neural network: the statistical substrate, frozen weights, no lifespan and no stake. Between them, what each has that the other cannot. {"id":"sl-hero","program":"akushnerphd-me","lane":"akushnerphd-me","subject":"substrate-law","type":"figure","created":"2026-07-27","title":"Wet vs dry - two substrates, permanently paired (v3 bigger type, clear weights)","provenance":{"session":"0c581a7f","tiles":"flux-dev via illustratassist; weighted-net inset hand-drawn"},"schema":1} THE WET SUBSTRATE chemical · lived-through-time stakes its outputs against a finite life w=.92 w=.71 w=.08 THE DRY SUBSTRATE statistical · frozen weights no lifespan, no time-stake ONLY THE WET HAS Presence · Recognition Judgment with stake · Authorship Refusal integrity under pressure ONLY THE DRY HAS Calculation at scale · Exhaustive search Synthesis across corpora Programmable uncertainty PERMANENTLY PAIRED different in kind, not in degree — the boundary does not move with scale

Philosophies and practical plans for socioeconomically responsible AI implementation — maximum diffusion at maximum safe speed, communicated to the public.

The Substrate Trilogy

The Substrate Law Read the paper →
Failing Legitimacy Read the paper →
The Transition Into the AI Age Read the paper →

Provenance in the AI Age

Task write APPLICATION X Any browser or local app Recording Workspace { chain: a3f9…c12e 7d2a…88bf } credential.json submit GitHub Timestamped commit Unalterable record Cloud record INSTITUTION / CORP ✓ verify the work how · when · with what help INDIVIDUAL ◎ rebut false accusations tamper-evident process log

AI made claiming authorship of generated work effortless — and after-the-fact detection is structurally impossible, producing failed catches and false accusations alike. The answer is provenance: git-backed hashing of individual input and output, recording how work was actually made.

  • Protects the honest: a tamper-evident record replaces guesswork — no failed detection, no irrefutable false accusation.
  • Generalizes past cheating: the same records make exploitative financial conduct identifiable fast (the 2008-crisis incentive inversion), instead of relying on moral courage against profit.
  • Gives digital backing to honest contribution accounting: “you get out of it what you put into it” — not punitive, honest — ending plausible-deniability free-riding on teams.

7–12 Writing Integrity (Ages 13+)

Demonstration Product: TeacherAware

AI made cheating effortless and detection impossible — failed catches and false accusations alike. TeacherAware records the writing process instead of guessing at the product: a digitally proctored workspace whose tamper-evident credential protects the honest student and answers the teacher in seconds. Free, LMS-friendly. Assign writing homework again!

Publishing Platform

The Citizen Scientist Exchange

Two panels comparing where peer-review capacity lives. Left, traditional publishing: nine authors all converge on two unpaid referees working on top of full research loads, and only a trickle reaches publication — review is scarce and shared. Right, the Citizen Scientist Exchange: eight endpoints ring a central hub, each endpoint carrying its own three frontier LLM reviewers, so review happens locally on the author's own keys for cents a round; the Exchange verifies everything and possesses nothing, and a human adjudicates escalation. Capacity scales with participation.

Verifiable AI-powered independent research — digitally peer reviewed by three frontier models, tamper-evident by construction, human-decided at the editorial desk. The publishing system for science at the speed of silicon, built with and powered by AI from bit 0. A "massive multiplayer" complement to institutional research, at a fraction of the cost.

Big-picture projects / startup seeds

The ideas are there and bootstrapped, but a real team and/or real time/funding are required to execute, deploy, maintain, and improve.

Education and the AI Transition

open-ed

“All books, learning materials, and assessments should be digital and interactive, tailored to each student and providing feedback in real time.”

Walter Isaacson, describing Jobs’ vision · Steve Jobs, 2011

“Gates sketched out his vision of what schools in the future would be like, with students watching lectures and video lessons on their own while using the classroom time for discussions and problem solving.”

Walter Isaacson, Steve Jobs · Bill Gates’ final visit to Jobs, 2011
Left: an aircraft solved in CFD, green streamlines sweeping past the wings. Right: two classrooms — students working at computer workstations, and students building small robots together at a lab bench.

Einstein had three kinds of luck: his mind, his moment, and the pricey school built for how he thought. The first two cannot be manufactured — the third now can, for anyone. Open-ed is adaptive, case-based, schema-before-facts, American-values-forward curriculum that meets each mind where it is, empowering teachers rather than replacing them.

The model

  • No labels, no tracks: natural strengths encouraged, gated on competence in the harder areas — eliminating the numerical-reasoning-as-intelligence tradeoff that leaves everyone else feeling inferior.
  • Proven structure (Alpha Schools): mastery-based AI academics in two hours a day, the rest for leadership, grit, financial literacy, entrepreneurship.
  • Pricing built for access: means-based for home-schoolers, for charter schools and public schools, within current public schools’ IT budgets.
  • The same engine retrains the worker crossing into the post-AI economy: a GI Bill at zero marginal cost, on infrastructure already in a billion pockets.

Live now

  • A full K-12 curriculum is a team project AI makes fast; meanwhile the case-based method targets adult scientific literacy — AI is happening to people, out of their control, and without a clear example of how it benefits them. The products below are built to change that. Two are live and free:

AI-Accelerated Bioengineering

The Bioengineering Machine

A flow diagram. A hallucinogen therapy idea leads to a laboratory proof, described as the credential. That proof and an existing general R&D AI workbench both feed into a specialised engine, the Bioengineering Machine, which in turn unlocks three application domains: existing drug and intermediate scale-up, biofuels and precursors, and synthetically challenging molecular and biomolecular libraries.

The flagship R&D Workbench, pointed at biology — with live success already on record.

  • De-risks a bioengineering program in silico — design, pathway, scale-up math — at roughly $10K where the wet-lab equivalent once demanded $10M.
  • First target: scalable hallucinogen-therapy products, behind a full-suite in-silico R&D engine.
  • Then the general targets the last biotech boom over-promised — biofuels among them — now armed with a decade of advances the original hype never had. One proof opens the floodgates.

Collaborators

Aaron Kushner

Aaron Kushner, Ph.D.

Principal · citizen scientist

LinkedIn →

Collaborations open

Scientific communications · AI product development · implementation consulting

Get in touch →

Click anywhere or press Esc to close