Software development. And no, that's not a joke. We really are roughly where we were in the 70's, we have not been able to make good use of the increase in machine power.
It would be nice to see something really fresh. I'm holding out for 'fleet', or something like it. A radical departure from what we think is 'best practice' and the ways to do things. Something that will turn software development into a true engineering discipline, without the house of cards feeling that we have today.
The idea here is to rebuild the existing software infrastructure that goes into a fully-functioning system such that the source code is at a minimum an order of magnitude smaller than systems are today. To do so, they claim, requires building the software stack on top of the right abstractions. So far, it seems they've made some excellent progress.
Software is only "technology" as painting once was. Much contemporary software is reminiscent of Medieval mosaics: big representations of highly conservative traditional models and crude depictions of people, built by a legion of anonymous workers, from monotonous fragile pieces. They feel like they were approached as engineering problems.
In Renaissance art, the new non-church wealthy elite still commissioned religious imagery, but they seem to've been after beauty more than direct reinforcement of their authority. This dynamic seems mirrored in the current divide between the products of startups and internal projects of enterprises.
It would be nice to see something really fresh. I'm holding out for 'fleet', or something like it. A radical departure from what we think is 'best practice' and the ways to do things. Something that will turn software development into a true engineering discipline, without the house of cards feeling that we have today.