THE VOIGHT-KAMPFF TEST FOR MAN, MACHINE, & BIG TECH.
DATA • CONVERSATIONS • SIGNALS • HUMANS • MACHINES
In 1982, Blade Runner imagined a future where the machines slipped their guardrails and a runner was sent to bring back the truth. That future arrived — not as fiction, but as an industry: trillion-dollar bets, machines that answer, power grids straining under the build. The film asked whether machines could pass for human. Our question is smaller, and harder: which claims about them can pass for true. Token Runner keeps the receipts — accountability for the ones that can, and the ones that can't. Read it. Question it. Walk it.
Source and method
Public vs Experts. Data: Pew Research, 2025, as published in Stanford HAI, AI Index Report 2026, figure 9.2.1. Values as published. Licensed CC BY-ND 4.0. The table and its date
Jobs: Public vs Experts. Data: Pew Research, 2025, as published in Stanford HAI, AI Index Report 2026, figure 9.2.4. Values as published. Licensed CC BY-ND 4.0. The table and its date
Trust in Government. Data: Ipsos, 2025, as published in Stanford HAI, AI Index Report 2026, figure 9.3.1. Values as published. Licensed CC BY-ND 4.0. The table and its date
Source and method
Debt by Year. Source: SEC XBRL companyfacts, 10-Q and 10-K cash-flow lines. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
CapEx vs Revenue. Source: SEC XBRL companyfacts, annual (10-K, fp FY, 340-400 day period). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Layoffs. Source: Layoffs.fyi, the AI layoffs list (https://layoffs.fyi/ai-layoffs/). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Source and method
Chip Owners, Chips by Designer. Source: Epoch AI, AI Chip Owners. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
CoWoS Wafers. Source: Epoch AI, AI Chip Components. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Source and method
China vs Rest. Source: Epoch AI, AI chip sales (https://epoch.ai/data/). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Export Rules. Source: Federal Register, Bureau of Industry and Security rules and proposed rules (https://www.federalregister.gov/). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Source and method
Accessibility. Source: Epoch AI, Notable AI Models. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
By Release Date, By Organisation. Source: Epoch AI, Frontier AI Models. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Source and method
Electricity Price. Source: U.S. Energy Information Administration, Electricity Data Browser / API v2, retail sales: average price, US, by sector. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Power by State, Price by State. Source: Epoch AI, the data centres file (https://epoch.ai/data/data_centers/data_centers.csv); U.S. Energy Information Administration for the state prices. Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
Source and method
Compute vs Capability, Open vs Closed. Source: Epoch AI, Epoch Capabilities Index (https://epoch.ai/data/). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
AI in Maths. Source: Epoch AI, AI acknowledgment trends in mathematics preprints (https://epoch.ai/data/). Adapted by Token Runner: filtered and added up for this chart. The table, its licence and its date
#7Demis HassabisCo-founder and chair of Google DeepMind; chief scientist of Alphabet
#12Yann LeCunFounder and executive chairman, AMI Labs; professor, NYU
#13Geoffrey HintonUniversity Professor Emeritus, University of Toronto
#29Yoshua BengioCo-president and scientific director, LawZero; professor, Université de Montréal
#49Chris OlahCo-founder and interpretability research lead, Anthropic