Save this video free

Mathematical Foundations of AI Alignment, Lecture 21 (AI Auditing and Forecast Testing)

Aaron Roth · 1:27:49 · Watch on YouTube

Part of University of Pennsylvania CIS 7000. See all 22 lectures and search them

These notes were made with YouTube Collector. Paste any YouTube link and get the same: a summary, chapters and a searchable transcript.

Overview

Aaron Roth formalizes auditing a forecasting model as distinguishing a forecaster that reports the true conditional probabilities from one that knows only the test. For any fixed finite deadline, if a test accepts every correctly informed forecaster with probability at least 1 minus α across all outcome distributions, the minimax theorem implies a test-aware randomized forecaster can pass with that probability on every outcome sequence. An unbounded-horizon tail test can evade that specific impossibility, but does not guarantee useful power on outcomes drawn from the real process; assumptions such as product-distributed outcomes can also change what is testable.

Key takeaways

Chapters

0:00 Auditing Whether Forecasts Reflect the True Data-Generating Process
3:10 Why Calibration Cannot Establish Understanding
5:44 Forecast Rules Can Be Queried on Counterfactual Histories
8:25 Defining an Informed Forecaster and the Audit Test
12:15 Uniform False-Negative Guarantees Set the Testing Standard
15:00 Why a Fixed Observation Deadline Matters
21:30 Handling Randomized Forecasting Algorithms
25:25 The Fixed-Horizon Impossibility Theorem
36:45 Turning Forecast Auditing into a Zero-Sum Game
40:30 Why the Finite-Horizon Game Has Finite Strategy Spaces
48:00 Minimax Converts an Informed Forecaster into a Test-Gaming Strategy
54:25 Removing the Fixed Deadline to Escape the Theorem
57:40 A Sequential Tail Test Based on the First Event
1:05:00 Why the Tail Test Controls False Negatives but Not Power
1:14:00 Unbounded Cutoffs Make the Order of Play Matter
1:20:00 Known Product Distributions Enable Stronger Forecast Checks
1:24:00 Efficient Test-Gaming and the Next Computational Question

Keep these chapters and the full searchable transcript in your own library.

Transcript excerpt

0:00 Right. I want to today think about an auditing or testing. Imagine that there is some phenomena that you would like to predict. You don't know the physical process underlying this phenomena. Think about I don't know, weather on Pluto or you know, you're trying to predict somehow relevant event. Point is like you know, what you're going to get to observe is a sequence of probabilistic forecasts things like you know, I think there's a

0:48 2% chance that the event happens today. Followed by realization of whether the event actually happened today or not. And you've got this forecasting technology that is feeding you these forecasts, but like what you're trying to audit for is whether this forecasting technology actually understands the underlying physical process governing probabilistic nature of these events. That is whether it is actually telling you the true conditional outcome probability given everything it has. Or whether instead it's not. It's somehow you know, doing something that is just trying to fool you into

1:36 thinking that it knows the underlying ground truth. So like on the one you're trying to sort of distinguish on the one hand like a fully informed forecaster that knows every day the true conditional label probabilities and tells them to you honestly. On the other hand, maybe like a charlatan, like a completely uninformed forecaster that knows the test it's trying pass, that knows how your auditing it. But otherwise um you know like knows nothing at all about the underlying physical process. And the question is like to what extent can we design tests that have some distinguishing power, that have some ability to tell the difference between like a forecasting technology that is correct versus one that is kind

2:27 of just trying to game the test. So um just to sort of

This is the start of the transcript. Save this video free for the whole transcript, searchable and time-coded.

In this course

Related notes

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Aaron Roth.

Want the full transcript?

Save this video in YouTube Collector to get its complete searchable transcript, your own AI summaries, and a library that keeps every video you collect in one place.