A post introduces 228 software tasks in its look at METR’s time horizons.
A post presents a statistical look back at METR’s time horizons and a look ahead, opening with 228 software tasks.
3.9K
2 Sources, first seen ago
—
Not ranked yet
—
Not ranked yet
@drewtnguyen# A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horizons chart, with human completion time (x) against estimated difficulty for AI (y). Does time proxy difficulty well? With @wfithian, I quantified this. 🧵
@JacobSteinhardtRT @drewtnguyen: # A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horiz…@drewtnguyen# A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horizons chart, with human completion time (x) against estimated difficulty for AI (y). Does time proxy difficulty well? With @wfithian, I quantified this. 🧵@JacobSteinhardtRT @drewtnguyen: # A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horiz…@drewtnguyen# A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horizons chart, with human completion time (x) against estimated difficulty for AI (y). Does time proxy difficulty well? With @wfithian, I quantified this. 🧵@JacobSteinhardtRT @drewtnguyen: # A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horiz…@drewtnguyen# A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horizons chart, with human completion time (x) against estimated difficulty for AI (y). Does time proxy difficulty well? With @wfithian, I quantified this. 🧵@JacobSteinhardtRT @drewtnguyen: # A statistical look back at METR’s time horizons—and a look ahead Here are the 228 software tasks from METR’s time horiz…
