Field Notes
The latest reading
What's actually moving the numbers, read in depth, with every claim cited. The companion to the live data, for when a number needs a thousand words.
Methodology · Oct 2026 · 5 min read
What these forecasts do not measure
The dated forecasts are not about AGI. Each one is about a single measurement in a single domain. What those numbers cover, and why raising the bar does not widen them.
The archive
Cited analysis on compute, cost, benchmarks, forecasts and resolved predictions. Filter by category.
Methodology
What these forecasts do not measure
The dated forecasts are not about AGI. Each one is about a single measurement in a single domain. What those numbers cover, and why raising the bar does not widen them.
Oct 2026 · 5 min read
Methodology
Why the forecast gives no single date
A point estimate hides everything that matters. The case for publishing a distribution, and how this one is built.
Oct 2026 · 4 min read
Data
Training compute is still doubling about every five months
The clearest leading indicator hasn't bent, and why it still does not move a single date on this site.
Oct 2026 · 4 min read
Analysis
Reading the benchmark plateau: saturation vs. real progress
When a 99% score stops meaning anything, and what this site does about it.
Oct 2026 · 4 min read
Forecast
The forecasters vs. the labs: where timelines diverge
Community forecasters and frontier researchers agree on direction and disagree on speed. Why the gap is shown here rather than averaged away.
Oct 2026 · 3 min read
Scoreboard
Scoring 2020's self-driving predictions, five years on
A cohort of confident calls came due. What resolving them teaches about reading hype.
Oct 2026 · 4 min read
Data
Cheaper is not the same as more capable
Falling cost is necessary, not sufficient. Where efficiency does and doesn't move the needle.
Oct 2026 · 3 min read
See the data
behind
the notes.
Ten domains, each on its own measure, with the gaps published beside the measurements.