Cross-posted from Bounded Regret.
Earlier this year, my research group commissioned 6 questions for professional forecasters to predict about AI. Broadly speaking, 2 were on geopolitical aspects of AI and 4 were on future capabilities:
Geopolitical: How much larger or smaller will the largest Chinese ML experiment be compared to the largest U.S. ML experiment, as measured by amount of compute used? How much computing power will have been used by the largest non-incumbent (OpenAI, Google, DeepMind, FB, Microsoft), non-Chinese organization? Future capabilities: What will SOTA (state-of-the-art accuracy) be on the MATH dataset? What will SOTA be on the Massive Multitask dataset (a broad measure of specialized subject knowledge, based on high school, college, and professional exams)? What will be the best adversarially robust accuracy on CIFAR-10? What will SOTA be on Something Something v2? (A video recognition dataset)
Forecasters output a probability distribution over outcomes for 2022, 2023, 2024, and 2025. They have financial incentives to produce accurate forecasts; the rewards total $5k per question ($30k total) and payoffs are (close to) a proper scoring rule, meaning forecasters are rewarded for outputting calibrated probabilities.
Depending on who you are, you might have any of several questions:
What the heck is a professional forecaster? Has this sort of thing been done before? What do the forecasts say? Why did we choose these questions? What lessons did we learn?
You’re in luck, because I’m going to answer each of these in the following sections! Feel free to skim to the ones that interest you the most.
And before going into detail, here were my biggest takeaways from doing this:
Projected progress on math and on broad specialized knowledge are both faster than I would have expected. I now expect more progress in AI over the next 4 years than I did previously. The relative dominance of
This article is purposely trimmed, please visit the source to read the full article.
The post Updates and Lessons from AI Forecasting appeared first on The Berkeley Artificial Intelligence Research Blog.