Public evidence function
The question outranks Founder 0, every program, and every preferred conclusion.
Parallel Human uses live venture work to study AI-native company formation. This record publishes the questions, methods, evidence, limits, and corrections. Frontier is the report series.
Frontier findings
Reports stay public, dated, and correctable. Email is for delivery and deeper artifacts, not a tollbooth around the evidence.
Evidence report
Software became abundant faster than evidence. What the first portfolio supports, what it does not, and what PH must test next.
Investor implication What capital is for if product formation gets cheaper.
Founder implication Which work agents absorb and which human constraints become more visible.
PH-SP-001 · Version 0.1 · Effective July 15, 2026
Each proposition is a Working hypothesis. The protocol defines its measures, thresholds, counterevidence, and conditions for revision or rejection. Read the public PH-SP-001 summary
Observed
Operator-supplied baseline, recorded July 13, 2026. These numbers describe one run. They are not presented as a universal benchmark.
Before this run: Russell did not have a software-development career. He did bring sales, strategy, operating experience, product judgment, and existing relationships.
Current finding: software output expanded faster than external evidence. Browse the portfolio or read the bounded report.
Current boundary
Founder 0 cannot establish repeatability, superior returns, or a universal venture count. Eleven public ventures do not mean eleven proven companies, and software output is not commercial evidence.
Report 001 is published with those limits intact. Missing costs, dates, decisions, and comparisons remain missing. They are not converted into a stronger conclusion.
Method
Each report connects an external claim to a real venture experiment, then checks it against dates, commits, deployments, spend, customer evidence, and known failures.
Founder 0 is one operator in one unusually concentrated run.
Reports include contradictory results and disclose missing records. They do not publish automated conclusions, invented market sizes, speculative exits, or fundraising claims without evidence and human approval.
For serious questions
Tell PH what you are trying to understand. Useful requests become manually reviewed relationship records in Attio. If a conversation can move the question forward, PH will send an invitation.
This is not an open calendar, investment allocation, securities offer, or LP commitment process.
For capital partners and serious observers
Monthly operating reports and evidence-triggered field notes about how one-person companies are changing. Follow cost to evidence, concentration decisions, failure, and capital timing. No recycled AI news.
No securities are being offered and no LP commitments are being accepted. This is the public research record.