Came across this old (2004) post from Moravec describing the evolution of his AGI timelines over time. Kudos to him, I say. Compute-based predictions seem to have historically outperformed every other AGI forecasting method (at least the ones that were actually used), as far as I can tell.
Came across this old (2004) post from Moravec describing the evolution of his AGI timelines over time. Kudos to him, I say. Compute-based predictions seem to have historically outperformed every other AGI forecasting method (at least the ones that were actually used), as far as I can tell.