Misuse is behavioral before it becomes a ticket.
Scenario thinking around manipulation, fraud, coercion, impersonation, escalation paths, evidence and the gap between written policy and real behavior.
Field experience for frontier technology teams
Human Systems
Human Systems
× AI
I help advanced technology teams understand what happens when powerful AI systems meet real people, institutions, vulnerable users, adversarial behavior and public expectations — then turn those realities into safer products, stronger deployment and measurable adoption.
I am not positioning myself as a substitute for an ML researcher. My value is the human operating environment around the model: behavior, safety, trust, institutions, youth and field execution.
01 / PROBLEM SPACE
Where technical capability meets human reality.
The strongest technical organizations still need people who can translate field behavior, institutional constraints, misuse patterns and public expectations into decisions teams can act on.
Safety and agency have to coexist.
Field-informed thinking about how young people actually adopt technology, where guardrails fail, and how access can increase capability without removing judgment.
Deployment fails when institutions cannot absorb it.
Translate between technology teams and municipalities, public bodies, educators, community organizations and other high-friction environments.
Bring the outside world into product decisions.
Design field loops with practitioners, users, civil society and researchers so product and policy teams see signals that conventional panels can miss.
02 / ראיות
A non-linear background. A coherent use.
The point is not the number of titles. It is the repeated exposure to people, risk, institutions and systems under real constraints.
FIELD BACKGROUNDCriminology → security → intelligence → municipal systemsExperience with behavior, risk, public responsibility and operational environments.See the sourced story ↗
YOUTHStartOnA youth-technology initiative built around access, belonging, mentors, media and opportunity rather than hardware alone.Open StartOn ↗
RESEARCH LAYERHuman agency, trust and responsible AIWorking frameworks that move from lived experience to questions, sources, critique and correction.Open research ↗
שרשרת המקור7YA evidence layerA source-linked public record designed to distinguish claims, dates, context and evidence instead of asking for blind trust.Inspect evidence ↗
03 / FIRST 90 DAYS
Start with a problem, not a title.
A practical engagement should create useful artifacts quickly. These are examples of the work I can own with the relevant internal experts.
Map the human risk surface
Interview internal teams and selected external users; identify misuse, trust and adoption failure modes; separate anecdotes from repeatable signals.
Build a field-evidence loop
Create a lightweight mechanism connecting product, policy, safety and outside practitioners so emerging problems חשיפה the right team with context.
Run one measurable pilot
Choose a bounded problem — youth-safe AI, public-sector adoption, user safety or AI literacy — define outcomes and produce a decision-ready evaluation.
04 / ROLE FAMILIES
Where this profile can create value.
AI Safety / SafeguardsTrust & SafetyResponsible AIProduct PolicyGlobal Affairs / PartnershipsUser & Field ResearchAI LiteracyPublic-Sector AdoptionSocial ImpactCross-functional Special Projects
05 / PRINCIPLE
Do not hire the biography. Use the perspective.
The proposition is deliberately testable: give me a real human-systems problem around AI, access to the relevant internal experts, and a bounded outcome. Judge the work on evidence, clarity and whether the organization makes a better decision afterward.
If your team owns the boundary between AI and society, we should talk.
Safety · trust · youth · public systems · adoption · research partnerships