What makes another couple big steps like that inevitable in a short time frame?

Before the recent floodgates cracked open, AI research made only slow incremental progress for decades. Why couldn't we already be back near that rate of progress?

Good point!