Just look at the progress of just about any large scale research project such as sequencing the human genome or making an AI capable of beating a chess grandmaster. The progress starts out unbearably slow, and then most of the gains happen in the last 20% of the time spent researching.
One reason for this is that as our technology increases, so does the power of the tools we use for creating more technology. Not only are computers getting better performance/price at an exponential rate, but those gains are fueling research into materials sciences, bioinformatics, and other emerging fields.
(Still it does say something about it, my country's population (small European country) is decreasing, around 10% less people in the last ten years even with all the immigration.)
My original point is that if the population is increasing at all, that will at least partially offset gains in efficiency brought about by newer technologies, and since population growth compounds over time, we'll need to make bigger and bigger gains just to maintain the status quo in terms of absolute energy and resource usage.