My life & career
Personal tools that once felt out of reach can become everyday helpers. The advantage shifts toward knowing how to ask good questions and evaluate answers.
THE YOTTABIT SCALE INDEX / AI economics and capability
286×
From $20 to seven cents.
A comparable level of AI capability became dramatically less expensive in under two years.
Nov 2022–Oct 2024 · Stanford AI Index 2025
THE YOTTABIT PERSPECTIVE
What does it mean to me?
Big change is fascinating. Its implications are what matter.
My business
If the software itself becomes widely available, differentiation comes from judgment, experience, workflow design and originality.
My industry
Finance, healthcare, manufacturing and education may see more tasks augmented, although regulated or high-stakes uses need real-world validation.
FROM JIM CARROLL’S WORK
This is the point behind Jim’s recent argument that when everyone has the same AI tools, the tools stop being the advantage.
Meet the futurist behind YottaBit ↗WHAT COULD I DO MONDAY MORNING?
Try one repetitive research or drafting task with an AI tool, then assess accuracy, cost and your own time saved.
THE EVIDENCE BEHIND THE STORY + METHODOLOGY & SOURCES
What the number really measures
Model inference price at fixed benchmark score, expressed as USD / million tokens.
The research foundation records: $20 to $0.07 per million tokens; Nov 2022-Oct 2024.
Where the claim has limits
Stanford benchmark-normalized example; non-identical model providers
Comparisons and historical rates must be independently checked against definitions and original measurements before an audited chart is published.