Direction B · Make multiplication visible
Find the missing gains.
The extra gains are easier to understand when you can see where they come from. Move both sliders.
Double the attempts. Double the success rate.
What happens when you do both?
What this model assumes
Expected successes = experiments per day × probability of success. The starting point is 10 × 10% = 1. If attempts improve by a factor A and the success rate by a factor B, the output improves by A × B. Adding only the separate gains gives 1 + (A − 1) + (B − 1). The missing interaction is (A − 1) × (B − 1).
Each cell is an expected value, not a guaranteed result. We assume the added experiments keep the chosen success rate and that all successful results can be used. This clean arithmetic explains why gains can multiply; it does not prove that every pair of AI scaling paths behaves this way. Improving one factor may change the other or cost more resources.
How this fits into a Substack post
Save the image above and add it to the post. Link the image, or a nearby “Try it yourself” link, to this page. The image also works in email. Substack does not support custom HTML or iframe code in its post editor. The compact version is for websites that allow iframes.