Loading Career Runway…
Loading Career Runway…
Track record
Dated, accountable predictions about how AI changes specific tasks. Every call resolves against the same graph that generates it. Accuracy is published whether it makes us look good or bad.
Accuracy summary
Every public Career Runway call is timestamped, evidence-graded, and resolves on a fixed deadline. We publish misses too — calibration matters more than vanity.
Total published
15
dated public calls
Resolved
0
0% of total
Correct direction
0
awaiting resolution
Avg resolution time
—
call_date → resolved_at
Grade distribution
Grades inherit from the underlying evidence band. Most calls are honestly C or D on the current substrate — we never upgrade confidence above the evidence that carries it.
Grade A
0%
n=0
Grade B
7%
n=1
Grade C
93%
n=14
Grade D
0%
n=0
Every published call, every resolution, and the reasoning behind it.
Calibration
Share of resolved calls that landed within their stated threshold. Partials and wrongs excluded from numerator. Withdrawn calls not counted.
Calibration data accumulates as calls resolve.
We show nothing until at least one prediction has been resolved. Honesty > optics.