AI Winter or Ice Age ended with AlexNext winning ImageNet competition in 2012 almost by 11%.
Google Trends show that environment has been heating up since then.
Are we now entering tropical times?
Just like in software, we made transition from science to engineering in 1980s/90s. Instead of implementing Bubble Sort or Quick Sort, using it effectively from some library became increasingly more important and useful. It was around that time that instead of Computer Scientists, the number of software engineers grew much more rapidly. (And in software, we are now in Technical era of commoditisation, where dev/hacking is a skill more common/affordable than robust engineering).
Similarly, in ML, are we moving from Scientists to Engineering era (and may be in a decade or two, will move to Technician era). Will now, more than understanding theory/algo, implementing robust engineering get more attention? Will the number of ML engineering grow more rapidly than Data Scientists?